*730Opinion
RICHARDSON, J.The most important of several problems which we consider is whether the principles of comparative negligence expressed by us in Li v. Yellow Cab Co. (1975) 13 Cal.3d 804 [119 Cal.Rptr. 858, 532 P.2d 1226, 78 A.L.R.3d 393], apply to actions founded on strict products liability. We will conclude that they do. We also inquire whether evidence of “compensating” safety devices installed in a motor vehicle by its manufacturer is admissible to offset alleged design deficiencies, and whether, under the particular facts herein, evidence of a driver’s claimed intoxication or of his asserted failure to use his vehicle’s safety equipment may be considered. While agreeing that evidence of compensating design characteristics is admissible, we will further determine that under the circumstances herein prejudicial error requiring reversal occurred upon the admission of evidence of the decedent’s alleged intoxication and failure to use safety devices in his vehicle.
The Facts And The Trial
Although there were no eyewitnesses, the parties agree, generally, on the reconstruction of the accident in question. In the early hours of October 31, 1970, decedent Kirk Daly, a 36-year-old attorney, was driving his Opel southbound on the Harbor Freeway in Los Angeles. The vehicle, while travelling at a speed of 50-70 miles per hour, collided with and damaged 50 feet of metal divider fence. After the initial impact between the left side of the vehicle and the fence the Opel spun counterclockwise, the driver’s door was thrown open, and Daly was forcibly ejected from the car and sustained fatal head injuries. It was equally undisputed that had the deceased remained in the Opel his injuries, in all probability, would have been relatively minor.
Plaintiffs, who are decedent’s widow and three surviving minor children, sued General Motors Corporation, Boulevard Buick, Underwriter’s Auto Leasing, and Aleo Leasing Company, the successive links in the Opel’s manufacturing and distribution chain. The sole theory of plaintiffs’ complaint was strict liability for damages allegedly caused by a defective product, namely, an improperly designed door latch claimed to have been activated by the impact. It was further asserted that, but for the faulty latch, decedent would have been restrained in the vehicle and, although perhaps injured, would not have been killed. Thus, the case involves a so-called “second collision” in which the “defect” did not *731contribute to the original impact, but only to the “enhancement” of injury.
At trial the jury heard conflicting expert versions as to the functioning of the latch mechanism during the accident. Plaintiffs’ principal witness testified that the Opel’s door was caused to open when the latch button on the exterior handle of the driver’s door was forcibly depressed by some protruding portion of the divider fence. It was his opinion that the exposed push button on the door constituted a design “defect” which caused injuries greatly in excess of those which Daly would otherwise have sustained. Plaintiffs also introduced evidence that other vehicular door latch designs used in production models of the same and prior years afforded substantially greater protection. Defendants’ experts countered with their opinions that the force of the impact was sufficiently strong that it would have caused the door to open resulting in Daly’s death even if the Opel had been equipped with door latches of the alternative designs suggested by plaintiffs.
Over plaintiffs’ objections, defendants were permitted to introduce evidence indicating that: (1) the Opel was equipped with a seat belt-shoulder harness system, and a door lock, either of which if used, it was contended, would have prevented Daly’s ejection from the vehicle; (2) Daly used neither the harness system nor the lock; (3) the 1970 Opel owner’s manual contained warnings that seat belts should be worn and doors locked when the car was in motion for “accident security”; and (4) Daly was intoxicated at the time of collision, which evidence the juiy was advised was admitted for the limited purpose of determining whether decedent had used the vehicle’s safety equipment. After relatively brief deliberations the jury returned a verdict favoring all defendants, and plaintiffs appeal from the ensuing adverse judgment.
Strict Products Liability And Comparative Fault
In response to plaintiffs’ assertion that the “intoxication-nonuse” evidence was improperly admitted, defendants contend that the deceased’s own conduct contributed to his death. Because plaintiffs’ case rests upon strict products liability based on improper design of the door latch and because defendants assert a failure in decedent’s conduct, namely, his alleged intoxication and nonuse of safety equipment, without which the accident and ensuing death would not have occurred, there is *732thereby posed the overriding issue in the case, should comparative principles apply in strict products liability actions?
It may be useful to refer briefly to certain highlights in the historical development of the two principles—strict and comparative liability. Tort law has evolved from a legal obligation initially imposed without “fault,” to recovery which, generally, was based on blameworthiness in a moral sense. For reasons of social policy and because of the unusual nature of defendants’ acts, liability without fault continued to be prescribed in a certain restricted area, for example, upon keepers of wild animals, or those who handled explosives or other dangerous substances, or who engaged in ultrahazardous activities. Simultaneously, and more particularly, those who were injured in the use of personal property were permitted recovery on a contract theory if they were the purchasers of the chattel or were in privity. Subsequently, liability was imposed in negligence upon the manufacturer of personalty in favor of the general consumer. (For a comprehensive historical review, see Prosser, Law of Torts (4th ed. 1971) § 96, pp. 641-644; 2 Harper & James, The Law of Torts (1956) § 12.2 and foll., p. 747 and foll.) Evolving social policies designed to protect the ultimate consumer soon prompted the extension of legal responsibility beyond negligence to express or implied warranty. Thus, in the area of food and drink a form of strict liability predicated upon warranty found wide acceptance. Warranty actions, however, contained their own inherent limitations requiring a precedent notice to the vendor of a breach of the warranty, and absolving him from loss if he had issued an adequate disclaimer.
General dissatisfaction continued with the conceptual limitations which traditional tort and contract doctrines placed upon the consumers and users of manufactured products, this at a time when mass production of an almost infinite variety of goods and products was responding to a myriad of ever-changing societal demands stimulated by wide-spread commercial advertising. From an historic combination of economic and sociological forces was bom the doctrine of strict liability in tort.
We, ourselves, were perhaps the first court to give the new principle judicial sanction. In Greenman v. Yuba Power Products, Inc. (1963) 59 Cal.2d 57 [27 Cal.Rptr. 697, 377 P.2d 897, 13 A.L.R.3d 1049], confronted with injuiy to an ultimate consumer caused by a defective power tool, we fastened strict liability on a manufacturer who placed on the market a defective product even though both privity and notice of breach of warranty were lacking. We rejected both contract and warranty theories, *733express or implied, as the basis for liability. Strict liability, we said, did not rest on a consensual foundation but, rather, on one created by law. The liability was created judicially because of the economic and social need for the protection of consumers in an increasingly complex and mechanized society, and because of the limitations in the negligence and warranty remedies. Our avowed purpose was “to insure that the costs of injuries resulting from defective products are borne by the manufacturer that put such products on the market rather than by the injured persons who are powerless to protect themselves.” (Id., at p. 63.) Subsequently, the Greenman principle was incorporated in section 402A of the Restatement Second of Torts, and adopted by a majority of American jurisdictions. (Prosser, supra, at pp. 657-658.)
From its inception, however, strict liability has never been, and is not now, absolute liability. As has been repeatedly expressed, under strict liability the manufacturer does not thereby become the insurer of the safety of the product’s user. (Cronin v. J. B. E. Olson Corp. (1972) 8 Cal.3d 121, 133 [104 Cal.Rptr. 433, 501 P.2d 1153]; Dippel v. Sciano (1967) 37 Wis.2d 443 [155 N.W.2d 55, 63]; West v. Caterpillar Tractor Company, Inc. (Fla. 1976) 336 So.2d 80, 90; Traynor, The Ways and Meanings of Defective Products and Strict Liability (1965) 32 Tenn.L.Rev. 363, 366-367.) On the contrary, the plaintiff’s injury must have been caused by a “defect” in the product. Thus the manufacturer is not deemed responsible when injury results from an unforeseeable use of its product. (Cronin, supra, at p. 126; Rest.2d Torts, supra, coms. g, h.) Furthermore, we have recognized that though most forms of contributory negligence do not constitute a defense to a strict products liability action, plaintiff’s negligence is a complete defense when it comprises assumption of risk. (Luque v. McLean (1972) 8 Cal.3d 136, 145 [104 Cal.Rptr. 443, 501 P.2d 1163]; Rest.2d Torts, § 402A, com. b.) As will thus be seen, the concept of strict products liability was created and shaped judicially. In its evolution, the doctrinal encumbrances of contract and warranty, and the traditional elements of negligence, were stripped from the remedy, and a new tort emerged which extended liability for defective product design and manufacture beyond negligence but short of absolute liability.
In Li v. Yellow Cab Co., supra, 13 Cal.3d 804, we introduced the other doctrine with which we are concerned, comparative negligence. We examined the history of contributory negligence, the massive criticism directed at it because its presence in the slightest degree completely barred plaintiff’s recovery, and the increasing defection from the doctrine. We then weighed the two principal arguments against its *734removal from California law, namely, that such a sharp change in direction required legislative action, and that there existed a cluster of asserted practical obstacles relating to multiple parties, the apportionment burdens on a jury and the uncertain effect on the defenses of last clear chance, assumption of risk, and wilful misconduct. Concluding that none of the obstacles was insurmountable, we announced in Li the adoption of a “pure” form of comparative negligence which, when present, reduced but did not prevent plaintiff’s recovery. (Pp. 828-829.) We held that the defense of assumption of risk, insofar as it is no more than a variant of contributory negligence, was merged into the assessment of liability in proportion to fault. (Pp. 824-825.) Within the broad guidelines therein announced, we left to trial courts discretion in the particular implementation of the new doctrine. (Pp. 826-827.)
We stand now at the point of confluence of these two conceptual streams, having been greatly assisted by the thoughtful analysis of the parties and the valuable assistance of numerous amici curiae. We are by no means the first to consider the interaction of these two developing principles.. As with the litigants before us, responsible and respected authorities have reached opposing conclusions stressing in various degrees the different considerations which we now examine.
Those counseling against the recognition of comparative fault principles in strict products liability cases vigorously stress, perhaps equally, not only the conceptual, but also the semantic difficulties incident to such a course. The task of merging the two concepts is said to be impossible, that “apples and oranges” cannot be compared, that “oil and water” do not mix, and that strict liability, which is not founded on negligence or fault, is inhospitable to comparative principles. The syllogism runs, contributory negligence was only a defense to negligence, comparative negligence only affects contributory negligence, therefore comparative negligence cannot be a defense to strict liability. (See Butaud v. Suburban Marine & Sport. Goods, Inc. (Alaska 1976) 555 P.2d 42, 47 (dis. opn. by Burke, J.), noted by Masin (1977) 4 Western St.U. L.Rev. 283, 284.) While fully recognizing the theoretical and semantic distinctions between the twin principles of strict products liability and traditional negligence, we think they can be blended or accommodated.
The inherent difficulty in the “apples and oranges” argument is its insistence on fixed and precise definitional treatment of legal concepts. In the evolving areas of both products liability and tort defenses, *735however, there has developed much conceptual overlapping and interweaving in order to attain substantial justice. The concept of strict liability itself, as we have noted, arose from dissatisfaction with the wooden formalisms of traditional tort and contract principles in order to protect the consumer of manufactured goods. Similarly, increasing social awareness of its harsh “all or nothing” consequences led us in Li to moderate the impact of traditional contributory negligence in order to accomplish a fairer and more balanced result. We acknowledged an intermixing of defenses of contributoiy negligence and assumption of risk and formally effected a type of merger, “As for assumption of risk, we have recognized in this state that this defense overlaps that of contributory negligence to some extent . . . .” (Li, supra, at p. 824.) In Li, we further reaffirmed our observation in Grey v. Fibreboard Paper Products Co. (1966) 65 Cal.2d 240, 245 [53 Cal.Rptr. 545, 418 P.2d 153]: “ ‘[T]hat in one kind of situation, to wit, where a plaintiff unreasonably undertakes to encounter a specific known risk imposed by a defendant’s negligence, plaintiff’s conduct, although he may encounter that risk in a prudent manner, is in reality a form of contributoiy negligence . . . .’ We think it clear that the adoption of a system of comparative negligence should entail the merger of the defense of assumption of risk into the general scheme of assessment of liability in proportion to fault in those particular cases in which the form of assumption of risk involved is no more than a variant of contributory negligence.” (13 Cal.3d at pp. 824-825, quoting Grey, supra, at pp. 245-246, italics in original.)
Furthermore, the “apples and oranges” argument may be conceptually suspect. It has been suggested that the term “contributory negligence,” one of the vital building blocks upon which much of the argument is based, may indeed itself be a misnomer since it lacks the first element of the classical negligence formula, namely, a duty of care owing to another. A highly respected torts authority, Dean William Prosser, has noted this fact by observing, “It is perhaps unfortunate that contributory negligence is called negligence at all. ‘Contributoiy fault’ would be a more descriptive term. Negligence as it is commonly understood is conduct which creates an undue risk of harm to others. Contributoiy negligence is conduct which involves an undue risk of harm to the actor himself. Negligence requires a duty, an obligation of conduct to another person. Contributory negligence involves no duty, unless we are to be so ingenious as to say that the plaintiff is under an obligation to protect the defendant against liability for the consequences of his own negligence.” (Prosser, Law of Torts, supra, § 65, p. 418.)
*736We think, accordingly, the conclusion may fairly be drawn that the terms “comparative negligence,” “contributory negligence” and “assumption of risk” do not, standing alone, lend themselves to the exact measurements of a micrometer-caliper, or to such precise definition as to divert us from otherwise strong and consistent counterváiling policy considerations. Fixed semantic consistency at this point is less important than the attainment of a just and equitable result. The interweaving of concept and terminology in this area suggests a judicial posture that is flexible rather than doctrinaire.
We pause at this point to observe that where, as here, a consumer or user sues the manufacturer or designer alone, technically, neither fault nor conduct is really compared functionally. The conduct of one party in combination with the product of another, or perhaps the placing of a defective article in the stream of projected and anticipated use, may produce the ultimate injury. In such a case, as in the situation before us, we think the term “equitable apportionment or allocation of loss” may be more descriptive than “comparative fault.”
Given all of the foregoing, we are, in the wake of Li, disinclined to resolve the important issue before us by the simple expedient of matching linguistic labels which have evolved either for convenience or by custom. Rather, we consider it more useful to examine the foundational reasons underlying the creation of strict products liability in California to ascertain whether the purposes of the doctrine would be defeated or diluted by adoption of comparative principles. We imposed strict liability against the manufacturer and in favor of the user or consumer in order to relieve injured consumers “from problems of proof inherent in pursuing negligence . . . and warranty . . . remedies, . . .” (Cronin v. J. B. E. Olson Corp., supra, 8 Cal.3d at p. 133, italics added; Greenman v. Yuba Power Products, Inc., supra, 59 Cal.2d at p. 63; Escola v. Coca Cola Bottling Co. (1944) 24 Cal.2d 453, 461-462 [150 P.2d 436] (conc. opn. by Traynor, J.).) As we have noted, we sought to place the burden of loss on manufacturers rather than “... injured persons who are powerless to protect themselves . . . .” (Greenman, supra, at p. 63; italics added; see Escola, supra, at p. 462; Price v. Shell Oil Co. (1970) 2 Cal.3d 245, 251 [85 Cal.Rptr. 178, 466 P.2d 722] [“protection of otherwise defenseless victims of manufacturing defects and the spreading throughout society of the cost of compensating them”] italics added.)
The foregoing goals, we think, will not be frustrated by the adoption of comparative principles. Plaintiffs will continue to be relieved of proving *737that the manufacturer or distributor was negligent in the production, design, or dissemination of the article in question. Defendant’s liability for injuries caused by a defective product remains strict. The principle of protecting the defenseless is likewise preserved, for plaintiff’s recovery will be reduced only to the extent that his own lack of reasonable care contributed to his injury. The cost of compensating the victim of a defective product, albeit proportionately reduced, remains on defendant manufacturer, and will, through him, be “spread among society.” However, we do not permit plaintiff’s own conduct relative to the product to escape unexamined, and as to that share of plaintiff’s damages which flows from his own fault we discern no reason of policy why it should, following Li, be borne by others. Such a result would directly contravene the principle announced in Li, that loss should be assessed equitably in proportion to fault.
We conclude, accordingly, that the expressed purposes which persuaded us in the first instance to adopt strict liability in California would not be thwarted were we to apply comparative principles. What would be forfeit is a degree of semantic symmetry. However, in this evolving area of tort law in which new remedies are judicially created, and old defenses judicially merged, impelled by strong considerations of equity and fairness we seek a larger synthesis. If a more just result follows from the expansion of comparative principles, we have no hesitancy in seeking it, mindful always that the fundamental and underlying purpose of Li was to promote the equitable allocation of loss among all parties legally responsible in proportion to their fault.
A second objection to the application of comparative principles in strict products liability cases is that a manufacturer’s incentive to produce safe products will thereby be reduced or removed. While we fully recognize this concern we think, for several reasons, that the problem is more shadow than substance. First, of course, the manufacturer cannot avoid its continuing liability for a defective product even when the plaintiff’s own conduct has contributed to his injury. The manufacturer’s liability, and therefore its incentive to avoid and correct product defects, remains; its exposure will be lessened only to the extent that the trier finds that the victim’s conduct contributed to his injury. Second, as a practical matter a manufacturer, in a particular case, cannot assume that the user of a defective product upon whom an injury is visited will be blameworthy. Doubtless, many users are free of fault, and a defect is at least as likely as not to be exposed by an entirely innocent plaintiff who will obtain full recovery. In such cases the manufacturer’s *738incentive toward safety both in design and production is wholly unaffected. Finally, we must observe that under the present law, which recognizes assumption of risk as a complete defense to products liability, the curious and cynical message is that it profits the manufacturer to make his product so defective that in the event of injury he can argue that the user had to be aware of its patent defects. To that extent the incentives are inverted. We conclude, accordingly, that no substantial or significant impairment of the safety incentives of defendants will occur by the adoption of comparative principles.
In passing, we note one important and felicitious result if we apply comparative principles to strict products liability. This arises from the fact that under present law when plaintiff sues in negligence his own contributory negligence, however denominated, may diminish but cannot wholly defeat his recovery. When he sues in strict products liability, however, his “assumption of risk” completely bars his recovery. Under Li, as we have noted, “assumption of risk” is merged into comparative principles. (13 Cal.3d at p. 825.) The consequence is that after Li in a negligence action, plaintiff’s conduct which amounts to “negligent” assumption of risk no longer defeats plaintiff’s recovery. Identical conduct, however, in a strict liability case acts as a complete bar under rules heretofore applicable. Thus, strict products liability, which was developed to free injured consumers from the constraints imposed by traditional negligence and warranty theories, places a consumer plaintiff in a worse position than would be the case were his claim founded on simple negligence. This, in turn, rewards adroit pleading and selection of theories. The application of comparative principles to strict liability obviates this bizarre anomaly by treating alike the defenses to both negligence and strict products liability actions. In each instance the defense, if established, will reduce but not bar plaintiff’s claim.
A third objection to the merger of strict liability and comparative fault focuses on the claim that, as a practical matter, triers of fact, particularly jurors, cannot assess, measure, or compare plaintiff’s negligence with defendant’s strict liability. We are unpersuaded by the argument and are convinced that jurors are able to undertake a fair apportionment of liability.
We are strengthened in the foregoing conclusion by the federal experience under the maritime doctrine of “unseaworthiness.” For decades, seamen have been permitted to recover from shipowners for injuries caused by defects rendering a vessel “unseaworthy.” (E.g., The *739Osceola (1903) 189 U.S. 158, 175 [47 L.Ed. 760, 764, 23 S.Ct. 483].) As noted by many courts, the concept of “unseaworthiness” is not limited to or affected by notions of the shipowner’s fault or due care, but applies to any deficiency of hull, equipment or crew, regardless of cause, which renders the ship less than reasonably fit for its intended purposes. (Mitchell v. Trawler Racer, Inc. (1960) 362 U.S. 539, 550 [4 L.Ed.2d 941, 948, 80 S.Ct. 926]; Seas Shipping Co. v. Sieracki (1946) 328 U.S. 85, 94 [90 L.Ed. 1099, 1105-1106, 66 S.Ct. 872], rehg. den., 328 U.S. 878 [90 L.Ed. 1646, 66 S.Ct. 1116].) Nonetheless, comparative principles have been made applicable to suits brought under the “unseaworthiness” doctrine, a form of strict liability, and the degree to which plaintiff’s own negligence contributes to his injuries has been considered in determining the amount of his recovery. (Pope & Talbot, Inc. v. Hawn (1953) 346 U.S. 406, 408-409 [98 L.Ed. 143,150-151,74 S.Ct. 202].) No serious practical difficulties appear to have arisen even where jury trials are involved. (E.g., Price v. Mosler (5th Cir. 1973) 483 F.2d 275, 277-278.)
We find equally unpersuasive a final objection that the merger of the two principles somehow will abolish or adversely affect the liability of such intermediate entities in the chain of distribution as retailers (Vandermark v. Ford Motor Co. (1964) 61 Cal.2d 256, 263 [37 Cal.Rptr. 896, 391 P.2d 168]), and bailors (Price v. Shell Oil Co., supra, 2 Cal.3d 245, 253.) We foresee no such consequence. Regardless of the identity of a particular defendant or of his position in the commercial chain the basis for his liability remains that he has marketed or distributed a defective product. If, as we believe, jurors are capable of assessing fully and fairly the legal responsibility of a manufacturer on a strict liability basis, no reason appears why they cannot do likewise with respect to subsequent distributors and vendors of the product.
We note that the majority of our sister states which have addressed the problem, either by statute or judicial decree, have extended comparative principles to strict products liability.
Our research discloses that of the more than 30 states which have adopted some form of comparative negligence, three (including California) have done so judicially. The two other states, Alaska (Butaud v. Suburban Marine & Sport. Goods, Inc., supra, 555 P.2d 42, 46) and Florida (West v. Caterpillar Tractor Company, Inc., supra, 336 So.2d 80, 89-90) have likewise, judicially, extended comparative principles to strict liability actions. At least five states have adopted comparative fault statutes which are not limited in their language to negligence actions: *740Arkansas (Ark.Stat.Ann. §§ 27-1763-1765 (1975)); Maine (Me.Rev.Stat., tit. 14, § 156 (1965)); Mississippi (3 Miss. Code Ann. § 11-7-15 (1917)); New York (CPLR, art. 14-A, § 1411 (1975)); and Rhode Island (2A R.I. Gen. Laws, § 9-20-4 (1971)). The New York statute expressly applies to strict liability actions. (See CPLR, supra, art. 14-A, § 1411, Practice Commentaries, CP 1411:1.) The Mississippi statute has been judicially construed as extending to suits founded on strict products liability. (Edwards v. Sears, Roebuck and Company (5th Cir. 1975) 512 F.2d 276, 290.) On the other hand one state, Connecticut, has statutorily prohibited the use of comparative fault as a defense in strict liability actions. (Pub.L. No. 77-335, eff. July 1, 1977.)
Of the three decisions which have declined to apply comparative negligence to strict liability, two have noted their reliance on state comparative negligence statutes which are expressly confined to “negligence” actions. (Melia v. Ford Motor Co. (8th Cir. 1976) 534 F.2d 795, 802 [Nebraska “slight-gross” comparative negligence statute]; Kirkland v. General Motors Corporation (Okla. 1974) 521 P.2d 1353, 1367-1368 [noting the limiting statutory language but holding that driving while intoxicated was product misuse barring recovery]; see also Kinard v. Coats Company, Inc. (1976) — Colo.App. — [553 P.2d 835, 837].) At least three jurisdictions have applied comparative negligence statutes to strict liability actions, despite language arguably limiting the statute application to negligence. (Dippel v. Sciano, supra, 155 N.W.2d 55, 64; Sun Val. Airlines, Inc. v. Avco-Lycoming Corp. (D.Idaho 1976) 411 F.Supp. 598, 602-603; Hagenbuch v. Snap-On Tools Corp. (D.N.H. 1972) 339 F.Supp. 676, 681-683.) Finally, one court has judicially extended a “pure” form of comparative fault to the traditional strict liability defense of “product misuse,” despite the existence of a statutory scheme of “modified” comparative negligence. (General Motors Corp. v. Hopkins (Tex. 1977) 548 S.W.2d 344, 351-352.)
Moreover, we are further encouraged in our decision herein by noting that the apparent majority of scholarly commentators has urged adoption of the rule which we announce herein. These include, from the academic community: Wade, A Uniform Comparative Fault Act—What Should It Provide? (1977) 10 Mich. J. L. Ref. 220; Fleming, The Supreme Court of California 1974-1975—Foreword: Comparative Negligence at Last—By Judicial Choice (1976) 64 Cal.L.Rev. 239, 269-271; Schwartz, Comparative Negligence (1974) § 12.1 et seq., p. 195 et seq. (see also Special Cal. Supplement re Nga Li v. Yellow Cab Co. of California (1975) § 4(B), p. 8); Wade, On the Nature of Strict Tort Liability for Products (1973) 44 *741Miss.L.J. 825, 850; Noel, Defective Products; Abnormal Use, Contributory Negligence, and Assumption of Risk (1972) 25 Vand.L.Rev. 93, 117-118; contra, Levine, Strict Products Liability and Comparative Negligence: The Collision of Fault and No-Fault (1977) 14 San Diego L.Rev. 337, 346 et seq. Among other commentaries urging such a rule are: Posner et al., Comparative Negligence in California: Some Legislative Solutions—Part II (1977) Los Angeles Daily Journal Report (Aug. 26, 1977) at pages 4, 9-18 (proposed legislation); Brewster, Comparative Negligence In Strict Liability Cases (1976) 42 J. Air L. & Com. 107, 109-117; Comment, Comparative Negligence in Vermont: A Solution or a Problem (1976) 40 Albany L.Rev. 777, 810; Feinberg, The Applicability of a Comparative Negligence Defense in a Strict Products Liability Suit Based on Section 402A of the Restatement of Torts 2d (1975) 42 Ins.Couns.J. 39, 52; Comment, Tort Defenses to Strict Products Liability (1969) 20 Syracuse L.Rev. 924, 925; Epstein, Products Liability: Defenses Based on Plaintiff’s Conduct (1968) 1968 Utah L.Rev. 267, 284; Levine, Buyer’s Conduct as Affecting the Extent of Manufacturer’s Liability in Warranty (1968) 52 Minn.L.Rev. 627, 652-663; contra, Robinson, Square Pegs (Products Liability) In Round Holes (Comparative Negligence) (1977) 52 State Bar J. 16; Schwartz, Pure Comparative Negligence in Action (1972) 34 Am.Trial Law. J. 117, 129.
. We find additional significance in the provisions of the proposed Uniform Comparative Fault Act (Act), authored by Professor Wade, a recognized torts scholar, distinguished professor of law, and former dean, Vanderbilt University, current reporter of the Restatement Second of Torts, and chairman of the special committee on the Act of the National Conference of the Commissioners on Uniform States Laws (Conference). Our attention has been called to the action of the Conference in August 1977, wherein it approved adoption of the Act by a vote of 40 states to 8 (California voting favorably). The Act is the distillation of approximately five years of discussion, analysis, and contribution by a special committee and a review committee of the Conference. We quote portions of section 1 of the proposed Act: “Section 1. [Effect of Contributory Fault.] [H] (a) In an action based on fault to recover damages for injury or death to person or harm to property, any contributory fault chargeable to the claimant diminishes proportionately the amount awarded as compensatory damages for an injury attributable to the claimant’s contributory fault, but does not bar recovery .... [If] (b) ‘Fault’ includes acts or omissions that are in any measure negligent or reckless toward the person or property of the actor or others, or that subject a person to strict tort liability.” (Italics added.) While lacking any legislative sanction, the *742Act, in our view, points in the direction of a responsible national trend. As such, section 1 is revealing in two notable respects: in its clear definitional expression in subsection (b) that comparative principles are to be applied to cases of “strict tort liability,” and in its substitution of the broad generic term “fault,” in subsection (a), as including both negligence and strict liability.
Having examined the principal objections and finding them not insurmountable, and persuaded by logic, justice, and fundamental fairness, we conclude that a system of comparative fault should be and it is hereby extended to actions founded on strict products liability. In such cases the separate defense of “assumption of risk,” to the extent that it is a form of contributory negligence, is abolished. While, as we have suggested, on the particular facts before us, the term “equitable apportionment of loss” is more accurately descriptive of the process, nonetheless, the term “comparative fault” has gained such wide acceptance by courts and in the literature that we adopt its use herein.
In Li, we announced a system of pure comparative negligence “the fundamental purpose of which shall be to assign responsibility and liability for damage in direct proportion to the amount of negligence of each of the parties.” (13 Cal.3d, at p. 829.) Those same underlying considerations of policy which moved us judicially in Li to rescue blameworthy plaintiffs from a 100-year-old sanction against all recovery persuade us now to extend similar principles to the strict products liability area. Legal responsibility is thereby shared. We think that apportioning tort liability is sound, logical and capable of wider application than to negligence cases alone. To hold otherwise, in our view, would be to perpetuate a system which, as we noted in Li, Dean Prosser describes as placing “. . . upon one party the entire burden of a loss for which two are, by hypothesis, responsible.” (Prosser, supra, § 67, p. 433.) We reiterate that our reason for extending a full system of comparative fault to strict products liability is because it is fair to do so. The law consistently seeks to elevate justice and equity above the exact contours of a mathematical equation. We are convinced that in merging the two principles what may be lost in symmetiy is more than gained in fundamental fairness.
In Li we scrupulously abstained from issuing a detailed guidebook to the new area of comparative negligence, preferring to adopt the view of a Florida court that “. . . the trial judges of this State are capable of applying [a] comparative negligence rule without our setting guidelines *743in anticipation of expected problems. The problems are more appropriately resolved at the trial level in a practical manner instead of a theoretical solution at appellate level. The trial judges are granted broad discretion in adopting such procedure as may accomplish the objectives and purposes expressed in this opinion.” (Hoffman v. Jones (Fla. 1973) 280 So.2d 431, 439-440.) We reaffirm the wisdom of such a course and, likewise, in the matter before us leave broad discretion in the trial court to implement the details of comparative principles in strict products liability cases.
For the guidance of trial courts, we do note the existence, under rule 49(a), Federal Rules of Civil Procedure, of a form of special verdict tailored to cases applying the maritime doctrine of strict liability for unseaworthiness, to which we have referred. Under this form, the jury is first required to answer “yes” or “no” to a series of questions setting forth possible bases for a finding that vessel unseaworthiness was a proximate cause of the plaintiff’s injuries. If the jury indicates that unseaworthiness was a contributing cause, it then moves on to a second group of similar questions seeking to determine whether, and in what particulars, the plaintiff’s own negligence was also a contributing factor. If the answers to these questions establish a finding of contributory negligence, the jury is told to “state in percentage the extent to which the plaintiff’s own negligence contributed to his injuries. (_%).” Finally, the juiy is instructed to indicate the amount of plaintiff’s damages without reference to his own negligence. The court then reduces the damage award by the percentage figure the jury has supplied. (3 Devitt & Blackmar, Federal Jury Practice and Instructions (3d ed. 1977) § 96.32; see also Menard v. Penrod Drilling Co. (5th Cir. 1976) 538 F.2d 1084, 1086-1087.) We cite this form as illustrative of one technique by which the court and jury may approach the task of apportionment.
Retroactivity
It remains for us to decide the extent to which comparative principles are to be applied to strict liability actions other than those hereafter filed. We conclude that, for reasons of public policy and the reasonable expectations of the parties to this action and litigants generally, the principles herein expressed shall apply to all cases in which trial has not begun before the date this opinion becojmes final in this court. No judgment based upon a trial which was commenced prior to the finality of this opinion shall be reversible on appeal on the sole ground that principles of comparative fault were not applied. If any such *744judgment is reversed on appeal for other reasons, the principles herein expressed shall be applicable to any retrial commenced after this opinion becomes final in this court. (Li v. Yellow Cab Co., supra, 13 Cal.3d at p. 829; Westbrook v. Mihaly (1970) 2 Cal.3d 765, 800 [87 Cal.Rptr. 839, 471 P.2d 487], vacated on other grounds (1971) 403 U.S. 915 [29 L.Ed.2d 692, 91 S.Ct. 2224].) As in Li, we give particular emphasis to “considerations of reliance applicable to individual cases according to the stage of litigation which they have reached ....” (Ibid.)
While it is arguable that retroactivity should extend to the finality of Li, and it is true that our conclusions herein owe a distinct philosophical debt to that case, nonetheless they do not constitute a direct application of Li. Furthermore, in Horn v. General Motors Corp. (1976) 17 Cal.3d 359, 370 [131 Cal.Rptr. 78, 551 P.2d 398], decided subsequent to Li, we retained, if only briefly, the separate system of defenses applicable to strict products liability. Under the circumstances before us, we conclude that it would be manifestly unfair to make the present opinion effective as of the finality of Li.
We further conclude that, under the particular circumstances, comparative principles cannot be applied retroactively to the instant case in order to justify admission of the intoxication and “nonuse” evidence here challenged. The issue of comparative fault was raised for the first time on appeal, and was never placed in issue by any party at trial herein. No jury instructions on the issue were requested or given. The juiy therefore had no basis for evaluating the evidence under correct principles of comparative fault. In the event of retrial, however, the principles herein announced will, of course, apply.
We turn to an explanation of the reasons which prompt us to reverse the judgment of the trial court, and then examine, for the benefit of court and counsel one of the remaining contentions of the parties.
The Safety Equipment And Intoxication Evidence
We must determine whether admission of evidence of decedent’s failure to use available safety devices and of his intoxication constituted prejudicial error under rules heretofore applicable to strict liability cases. We conclude that it did.
*745While initially evidence bearing on decedent’s intoxication was excluded, other evidence pertaining to the decedent’s alleged failure to employ seat belts and door locks was admitted, apparently on the ground that nonuse of safety devices bore on the issues of proximate cause and mitigation of damages. Plaintiffs contended that evidence of Daly’s intoxication, or of his failure to use available safety devices, was wholly inadmissible since contributory negligence was not a defense to an action founded in strict liability for a defective product. (Horn v. General Motors Corp., supra, 17 Cal.3d at pp. 369-371; Luque v. McLean, supra, 8 Cal.3d 136, 145.) The trial court ultimately admitted the intoxication evidence, ruling that such evidence related to decedent’s failure to use the Opel’s safety devices, which failure, the court reasoned, would bar recovery on the theory of product misuse “aside from any question of contributory negligence.” (In fairness to the very able trial judge, it must be noted that the trial herein preceded rendition of our opinions in both Li and Horn.)
As we observed in Horn, to accept a “nonuse” of safety equipment as a complete defense to a products liability action would constitute but a thinly disguised subversion of the rule that contributory negligence does not prevent recovery. Horn expressly rejected arguments that such “nonuse” could defeat recovery on theories of “assumption of risk,” “product misuse,” “proximate cause,” or “mitigation of damages.” (Id., at pp. 369-371.)
Substantial time was spent on the nonuse and intoxication issues, reasonably suggesting to tíre jury their central importance to the defense case. There can be little doubt that the evidence of Daly’s intoxication was inflammatory. The only restrictions placed on the jury’s consideration of the intoxication evidence was that it bore on the “nonuse” of safety devices in general. No limitation was placed on the conclusions which the jury could draw either from a finding of “nonuse” itself, or as to the effect on its deliberations of a finding of “nonuse.” In the absence of any such restrictions, we think the jury could well have concluded that decedent’s negligent failure, induced by intoxication, to use the belts and locks constituted negligent conduct which completely barred recovery for his death. We do not think it reasonable to conclude that plaintiffs waived their objection by failing to request limiting instructions.
In summary, our review of the record convinces us that, notwithstanding that plaintiffs’ case was founded on strict products liability, evidence of decedent’s failure to use available seat belts and door locks, and of his *746intoxication at the time of the fatal collision, may have been improperly regarded by the jury as authorizing a defense verdict. It appears reasonably probable that, had such evidence been either excluded or its effect confined, a result more favorable to plaintiffs would have been reached. Reversal is therefore required. (People v. Watson (1956) 46 Cal.2d 818, 836 [299 P.2d 243]; Downing v. Barrett Mobile Home Transport, Inc. (1974) 38 Cal.App.3d 519, 525 [113 Cal.Rptr. 277]; see Henderson v. Harnischfeger Corp. (1974) 12 Cal.3d 663, 670-676 [117 Cal.Rptr. 1, 527 P.2d 353].)
Defect—Component Or Product As A Whole?
We examine for the benefit of court and counsel, in the event of retrial, a single remaining contention of plaintiffs.
Plaintiffs challenge a jury instruction which directed that “[i]n determining whether or not the vehicle was defective you should consider all of the equipment on the vehicle including any features intended for the safety of the driver.” They urge that only the precise malfunctioning component itself, and alone, may be considered in determining whether injury was caused by a defectively designed product. We disagree, concluding that the issue of defective design is to be determined with respect to the product as a whole, and that the trial court’s instruction was correct.
The jury could properly determine whether the Opel’s overall design, including safety features provided in the vehicle, made it “crashworthy,” thus rendering the vehicle nondefective. Product designs do not evolve in a vacuum, but must reflect the realities of the market place, kitchen, highway, and shop. Similarly, a product’s components are not developed in isolation, but as part of an integrated and interrelated whole. Recognizing that finished products must incorporate and balance safety, utility, competitive merit, and practicality under a multitude of intended and foreseeable uses, courts have struggled to evolve realistic tests for defective design which give weight to this necessary balancing. Thus, a number of California cases have recognized the need to “weigh” competing considerations in an overall product design, in order to determine whether the design was “defective.” Recently, we ourselves in Barker v. Lull (1978) ante, pp. 413, 431 [143 Cal.Rptr. 225, 573 P.2d 443], have described some of the factors to be considered. (See, also, Buccery v. General Motors Corp. (1976) 60 Cal.App.3d 533, 547 [132 Cal.Rptr. 605]; Baker v. Chrysler Corp. (1976) 55 Cal.App.3d 710, 716 [127 Cal.Rptr. 745]; Hyman v. Gordon (1973) 35 Cal.App.3d 769, 773 [111 Cal.Rptr. 262].)
*747The danger of piecemeal consideration of isolated components has been expressly recognized. (Self v. General Motors Corp. (1974) 42 Cal.App.3d 1, 6-7 [116 Cal.Rptr. 575]; Dreisonstok v. Volkswagenwerk A.G. (4th Cir. 1974) 489 F.2d 1066, 1071-1072, applying Virginia law.) Specifically, it has been observed that a design rendered safe in one situation may become more dangerous in others. (Self supra, at pp. 6-7.) However phrased, these decisions emphasize the need to consider the product as an integrated whole.
We find that plaintiffs’ other contentions lack merit.
Conclusion
It is readily apparent that the foregoing broad expressions of principle do not establish the duties of the jury with that fixed precision which appeals to minds trained in law and logic. Nonetheless, rather than attempt to anticipate every variant and nuance of circumstance and party that may invoke comparative principles in a strict products liability context, we deem it wiser to await a case-by-case evolution in the application of the broad principles herein expressed.
By extending and tailoring the comparative principles announced in Li, supra, to the doctrine of strict products liability, we believe that we move closer to the goal of the equitable allocation of legal responsibility for personal injuries. We do so by relying on what Professor Schwartz aptly terms a “predicate of fairness.” In making liability more commensurate with fault we undermine neither the theories nor the policies of the strict liability rule. In Li we took “a first step in what we deem to be a proper and just direction, . . .” (13 Cal.3d at p. 826.) We are convinced that the principles herein announced constitute the next appropriate and logical step in the same direction.
The judgment is reversed.
Tobriner, J., Clark, J., and Manuel, J., concurred.
CLARK, J.The reasoning of Li v. Yellow Cab Co. (1975) 13 Cal.3d 804 [119 Cal.Rptr. 858, 532 P.2d 1226, 78 A.L.R.3d 393], as the majority point out, is equally applicable to strict liability cases and compels applying *748comparative fault in those cases. (Horn v. General Motors Corp. (1976) 17 Cal.3d 359, 372 [131 Cal.Rptr. 78, 551 P.2d 398] (dis. opn. of Clark, J.).) Under the compulsion of Li, I have signed the majority opinion.
Nevertheless, again we must recognize the difficulties inherent in comparing fault. (See American Motorcycle Assn. v. Superior Court (1978) ante, pp. 578, 608 [146 Cal.Rptr. 182, 578 P.2d 899] (dis. opn. of Clark, J.).) Relying on the apples and oranges argument, Justices Mosk and Jefferson point out that comparative fault cannot be applied logically and consistently in strict liability cases. (Infra, pp. 762-764, 751 et seq.) The difficulty, however, isnotlimitedto comparingstrictliabilitywith negligence. The same difficulty persists in almost every case in which we attempt to compare parties’ negligence. (Horn v. General Motors Corp., supra, 17 Cal.3d 359, 377 (dis. opn.).)
For example, assume three drivers collide at an intersection, one being intoxicated, the second speeding, and the third having gone through a stop signal. Neither logic nor common experience can tell us how much of the loss is attributable to each driver. Logic cannot tell us that the first should bear 10 percent of the loss, the second 30, and the third 60.1 Nor may it establish any other percentages of liability. When we substitute a defective product for one of the drivers’ negligence, we neither add to nor subtract from the difficulty of the comparison, and the problem of comparison is the same whether we compare different negligent acts or compare a negligent act to a defect.
Logic failing, is there a hope that juries, or judges when juries are waived, will arrive at consistent results? Obviously not.2 It is not a problem that the allocation of fault cannot be precisely measured —rather, in most cases there is no measuring standard.
The lack of logic and consistency not only means that most claims will not be disposed of equitably but also that attorneys may seldom reasonably evaluate the cases for purposes of settlement. And because settlement plays such a large part in the determination of accident claims the efficient administration of justice is substantially impaired.
*749I also must part company from Justices Mosk and Jefferson as to the effect of the apples and oranges argument. The difficulty in comparing fault does not warrant either returning to the all-or-nothing rule applicable to contributory negligence cases prior to Li or continuing the all-or-nothing rule applied in strict liability cases. Dean Prosser’s famous statement remains unanswerable. It is improper to place “upon one party the entire burden of a loss for which two are, by hypothesis, responsible.” (Prosser, Torts (4th ed. 1971) § 67, p. 433.) The statement is as applicable to negligence cases as it is to strict liability cases. Apples and oranges do not warrant denial of loss apportionment—they require establishing a better system of apportionment.
Li effectively pointed out that the existing contributory negligence system placed on one party the entire burden of a loss for which two were responsible (13 Cal.3d at p. 810, fn. 3), and today’s majority opinion effectively points out that the negligent plaintiff is responsible and should not recover as much as an innocent one. (Ante, pp. 736-737.)
Those principles do not require a comparative fault system. Can they not be satisfied by a system which establishes a uniform index factor, such as 30, 50 or 70 percent? A uniform discount of the negligent plaintiff’s recovery would eliminate the necessity of the often impossible task of comparing fault. A discount system would bring about consistency and predictability where neither now exists, permitting evaluation and settlement of claims.
Such an approach is consistent with the departure from the Li principle made by the majority in American Motorcycle Assn. v. Superior Court, supra, ante, page 578, and in the instant case. In Li, this court repeatedly emphasized that the extent of fault should govern the extent of “liability” (e.g., 13 Cal.3d at pp. 811, 813, 829). By repeated use of the word “liability” rather than recovery the emphasis was placed on fairness to the defendant in the new comparative fault doctrine, and throughout the opinion comparability of fault is emphasized. However, a selective rejection of the Li reasoning appears in American Motorcycle in connection with the discussion of joint and several liability. The faimess-to-defendant justification for comparative fault is ignored, rather the court relies upon a policy of compensating injured plaintiffs. (Ante, p. 578.) And the court suggests that there is a significant difference between plaintiff negligence—a failure to exercise due care for oneself —and defendant negligence—a failure to exercise due care for others. (Ante, p. 578.) Today’s opinion, telling us that “the ‘apples and oranges’ *750argument may be conceptually suspect,” similarly emphasizes the difference between plaintiff and defendant fault, pointing out the difficulty in finding a breach of duty upon which to predicate plaintiff negligence. (Ante, pp. 735-736.)
In cases where the plaintiff is not negligent, the determination of liability is based on fault, and the extent of liability is based solely on amount of loss—whether grossly or marginally negligent, the defendant’s liability is the same. By focusing on a policy of compensating injured plaintiffs, pointing out the difference between plaintiff and defendant negligence, and providing that plaintiff’s recovery is to be diminished on the basis of his fault without regard to defense fairness, this court has departed from the Li principle, reflecting recognition of its shortcomings. We also should recognize that comparative fault—although equitable in theory—cannot be applied equitably, and precludes consistency of result, making settlement much more difficult and substantially impairing the efficient administration of justice.
We can and should do a better job. A discount system will not eliminate all inequities. Arbitrary results are inherent in attempting to adjust a loss between the two or more parties responsible, when their fault is based on different acts. The present comparative system is not only inequitable and arbitrary but also inconsistent and unpredictable. Eliminating two of the four defects by adoption of a discount system is worthy of the court’s task.
Having úndertaken the legislative function by repudiating contributory negligence and adopting comparative fault, we have abandoned our traditional deference to legislative province and to stare decisis. Having done so, we, like the Legislature, should reconsider our bold decisions from time to time, performing the legislative process to the best of our ability—until the Legislature awakens to reclaim and exercise its historic power.
There will be a few situations where logic may be applicable, but even in those cases, it will not be very helpful. If both drivers causing an accident were speeding, the one driving at the greater speed should bear the greater responsibility. But is the greater responsibility 51 percent of the loss or 99 percent of the loss?
In the instant case, plaintiff claims there was a defective door lock. Defendant claims plaintiff was intoxicated and failed to fasten seatbelts. Assuming both parties prove their claims, I could not find fault with either a 5 percent recovery or a 95 percent recovery.