“[T]he problem [of negligent infliction of emotional distress] should be solved by the application of the principles of tort, not by the creation of exceptions to them. Legal history shows that artificial islands of exceptions, created from the fear that the legal process will not work, usually do not withstand the waves of reality and, in time, descend into oblivion.” (Dillon v. Legg (1968) 68 Cal.2d 728, 747 [69 Cal.Rptr. 72, 441 P.2d 912, 29 A.L.R.3d 1316] [Tobriner, J., admonishing this court not to do what the majority do today].)
The majority grope for a “bright line” rule for negligent infliction of emotional distress actions, only to grasp an admittedly arbitrary line which will deny recovery to victims whose injuries from the negligent acts of others are very real. In so doing, the majority reveal a myopic reading of Dillon v. Legg, supra, 68 Cal.2d 728. They impose a strict requirement that plaintiff be present at the scene of the injury-producing event at the time it occurs and is aware that it is causing injury to the victim. This strict requirement rigidities what Dillon forcefully told us should be a flexible rule, and will lead to arbitrary results. I would follow the mandate of Dillon and maintain that forseeability and duty determine liability, with a view toward a policy favoring reasonable limitations on liability. There is no reason why these general rules of tort law should not apply to negligent infliction of emotional distress actions.
I.
We held in Dillon that a mother who witnesses the negligent infliction of death or injury on her child may recover for the resulting emotional distress even though the mother does not fear imminent physical harm. We recognized that the primary consideration in finding liability was foreseeability (Dillon, supra, 68 Cal.2d at pp. 730-740) and rejected the “hopeless artificiality” of the zone-of-danger rule. (Id. at p. 733.)
The majority themselves note that “foreseeability of the injury [is] the basis of a negligent actor’s duty,” (maj. opn. at p. 654) and quote from Dillon that this issue must necessarily be adjudicated only upon “a case-by-case basis.” (Maj. opn. at p. 655, quoting Dillon, supra, 68 Cal.2d at p. 741.) “ ‘We cannot now predetermine defendant’s obligation in every situation by *683a fixed category; no immutable rule can establish the extent of that obligation for every circumstance of the future.’ ” (Maj. opn. at p. 654, quoting Dillon, supra, 68 Cal.2d at p. 740, italics added.)
Though Dillon made foreseeability its lodestar, it provided three factors for courts to consider in determining whether a negligent infliction of emotional distress cause of action was stated in a particular case: “(1) Whether plaintiff was located near the scene of the accident as contrasted with one who was a distance away from it. (2) Whether the shock resulted from a direct emotional impact upon plaintiff from the sensory and contemporaneous observance of the accident, as contrasted with learning of the accident from others after its occurrence. (3) Whether plaintiff and the victim were closely related, as contrasted with an absence of any relationship or the presence of only a distant relationship.” (Dillon, supra, 68 Cal.2d at pp. 740-741.)
Dillon denounced “artificial abstractions which bar recovery contrary to the general rules” of tort law, and emphasized that “ ‘mechanical rules of thumb which are at variance with these principles do more harm than good.’ [Citation.]” (68 Cal.2d at pp. 746-747.) However, some courts have rigidly and mechanically applied the Dillon guidelines—many times at the cost of injustice to a victim of a tortfeasor’s negligent act. Professor Diamond, although quoted extensively by the majority, warned against a strict application of the Dillon guidelines: “[C]ourts have applied the Dillon guidelines mechanically, viewing them as strict preconditions to recovery. This mechanical application has led to the erection of arbitrary limitations on recovery bearing little relation to the principles of foreseeability espoused so forcefully in Dillon. While in some instances mental distress is compensated, other equally foreseeable mental injuries are not. The result is feast or famine for the plaintiff depending on the fortuities of time, location, or characterization of the plaintiff as ‘direct’ or ‘indirect.’ ” (Diamond, Dillon v. Legg Revisited: Toward a Unified Theory of Compensating Bystanders and Relatives for Intangible Injuries (1984) 35 Hastings L.J. 477, 477-478.)
The majority ignore the fundamental mandate of Dillon to consider foreseeability and duty in finding liability. Their only justification for this and a strict rule that will limit liability at the cost of arbitrary results is an amorphous “policy” one. They ironically use the term “amorphous” to describe the concepts of foreseeability and duty and state that “[i]n the present context, however, we are concerned not with whether an injury is ‘foreseeable’ as a result of the negligent conduct. . . . [T]he court’s role in deciding whether a ‘duty’ to these persons should be recognized does not depend solely on the ‘foreseeability’ of the emotional distress, but on these policy considerations.” (Maj. opn. at p. 654, fn. 3, italics added.) “[I]t is clear that *684foreseeability of the injury alone is not a useful ‘guideline’ or a meaningful restriction on the scope of the NIED action. . . . [fl] [Pjolicy considerations justify restrictions on recovery for emotional distress notwithstanding the sometimes arbitrary result, and that the court has an obligation to establish those restrictions.” (Maj. opn. at pp. 663-664, italics added.) The majority admit their “policy” reasons are only a balance of “arbitrary lines which deny recovery to some victims whose injury is very real against that of imposing liability out of proportion to culpability for negligent acts,” with a view to the “administration of justice of clear guidelines.” (Id. at p. 664.)
For these reasons, the majority impose the strict requirement that plaintiff be present at the scene of the injury producing event at the time it occurs and is then aware that it is causing injury to the victim.1 They freely admit to “drawing arbitrary lines” but complain that it is “unavoidable if we are to limit liability and establish meaningful rules for application by litigants and lower courts.” (Maj. opn. at p. 666.) Thus what in Dillon were guidelines to assist courts in assessing liability become a tripartite test, which includes the above-mentioned strict and arbitrary requirement, and displaces the consideration of foreseeability.
Under the majority’s strict requirement, a mother who arrives moments after an accident caused by another’s negligence will not be permitted recovery. No matter that the mother would see her six-year-old son immediately after he was electrocuted, lying in a puddle of water in a dying state, gagging and choking in his own vomit, as in Hathaway v. Superior Court (1980) 112 Cal.App.3d 728 [169 Cal.Rptr. 435], No matter that the mother would be following her daughters’ car and would come upon the wreckage of the car “before the dust had settled” to find the mangled bodies of her daughters, who were dead or dying, as in Parsons v. Superior Court (1978) 81 Cal.App.3d 506, 509 [146 Cal.Rptr. 495],
The answer to the question of how this court should limit liability does not lie in the majority’s rigid application of Dillon and the toleration of arbitrary results that will flow therefrom. As the Wyoming Supreme Court suggested, in Gates v. Richardson (Wyo. 1986) 719 P.2d 193, the nature of the shock to be compensated requires a realistic approach to the contempo*685raneous-perception factor: “It is more than the shock one suffers when he learns of the death or injury of a child, sibling or parent over the phone, from a witness, or at the hospital. It is more than bad news. The kind of shock the tort requires is the result of the immediate aftermath of an accident. It may be the crushed body, the bleeding, the cries of pain, and in some cases, the dying words which are really a continuation of the event. The immediate aftermath may be more shocking than the actual impact. . . .” (Gates, supra, at p. 199; see also Dziokonski v. Babineau (1978) 375 Mass. 555 [380 N.E.2d 1295, 1302]; Comment, Dillon Revisited: Toward a Better Paradigm for Bystander Cases (1982) 43 Ohio St. L.J. 931, 948.) The court there held that a close relative generally should be permitted recovery if she “observed the serious bodily harm or death shortly after its occurrence but without material change in the condition and location of the victim.” (Gates, supra, 719 P.2d at p. 199; see also Portee v. Jaffee (1980) 84 N.J. 88 [417 A.2d 521] [recovery allowed to mother who arrived minutes after her son became wedged in an elevator shaft and was present and listened to his moans and cries during the futile four-and-one-half-hour struggle to free him before he died].)
The majority’s strict requirement does not simply comprise a “bright line” rule that rationally limits liability. It is arbitrary and will lead to unjust results. Dillon condemned the “hopeless artificiality” that the majority propounds, and decried the “artificial abstractions which bar recovery contrary to the general rules” of tort law. The requirement is exactly the “mechanical rules of thumb” that Dillon explicitly admonished us not to create. We should follow Dillon and its progeny and maintain the rational and traditional rule that reasonable foreseeability is the basis for determining liability. (68 Cal.2d at p. 740.) “In order to limit the otherwise potentially infinite liability which would follow every negligent act, the law of torts holds defendant amenable only for injuries to others which to defendant at the time were reasonably forseeable.” (Id. at p. 739, italics added.) As Dillon instructed, there is “no good reason why the general rules of tort law, including the concepts of negligence, proximate cause, and foreseeability, long applied to all other types of injury, should not govern the case now before us.” (Id. at p. 746.) Dillon's test of reasonable foreseeability “facilitates rational risk spreading and correlates liability with the risks that the defendant should expect.” (Diamond, supra, 35 Hastings L.J. at p. 500.)
II.
Of course I share the majority’s policy concern that tortfeasors not face unlimited liability for their negligent acts. As stated above, the Dillon court recognized foreseeability as a general limit on tort liability. The court stated that the purpose of the three guidelines was actually to limit a defendant’s *686liability to “injuries to others which to defendant at the time were reasonably foreseeable.” (Dillon, supra, 68 Cal.2d at p. 739; see Rabin, Tort Recovery for Negligently Inflicted Economic Loss: A Reassessment (1985) 37 Stan.L.Rev. 1513, 1524-1526.)
Although I disagree with the majority’s method of placing undue and what appears to be almost total reliance on a policy rationale, the Dillon guidelines also attempt to implement public policy in favor of reasonable limitations on liability. (See Ochoa v. Superior Court (1985) 39 Cal.3d 159, 180-181 [216 Cal.Rptr. 661, 703 P.2d 1] (Grodin, J., conc.); see also Pearson, Liability to Bystanders for Negligently Inflicted Emotional Harm—A Comment on the Nature of Arbitrary Rules (1982) 34 U.Fla.L.Rev. 477, 505-506.) The Dillon court did not intend the guidelines to be exclusive (68 Cal.2d at p. 740), and it specifically reserved the question of “whether, in the absence or reduced weight of some of the [three listed] factors, we would conclude that the accident and injury were not reasonably foreseeable and that therefore defendant owed no duty of care to plaintiff.” (Id. at p. 741).
To determine whether defendants owed plaintiff a duty of care in this case, I think it is fruitful to reexamine the second Dillon guideline in light of Rowland v. Christian (1968) 69 Cal.2d 108 [70 Cal.Rptr. 97, 443 P.2d 561, 32 A.L.R.3d 496], our leading case defining a defendant’s duty of care. In Rowland, decided just two months after Dillon, we held that in the absence of a statutory exception to the legislative mandate that all persons are liable for injuries caused by failure to exercise due care (Civ. Code, § 1714, subd. (a)), “no such exception should be made unless clearly supported by public policy. [Citations.] [1|] A departure from this fundamental principle involves the balancing of a number of considerations; the major ones are the foreseeability of harm to the plaintiff, the degree of certainty that the plaintiff suffered injury, the closeness of the connection between the defendant’s conduct and the injury suffered, the moral blame attached to the defendant’s conduct, the policy of preventing future harm, the extent of the burden to the defendant and consequences to the community of imposing a duty to exercise care with resulting liability for breach, and the availability, cost, and prevalence of insurance for the risk involved.” (Rowland, supra, 69 Cal.2d at pp. 112-113; see also Sun'n Sand, Inc. v. United California Bank (1978) 21 Cal.3d 671, 695 [148 Cal.Rptr. 329, 582 P.2d 920]; Tarasoff v. Regents of University of California (1976) 17 Cal.3d 425, 434-435 [131 Cal.Rptr. 14, 551 P.2d 334, 83 A.L.R.3d 1166]; Weirum v. RKO General, Inc. (1975) 15 Cal.3d 40, 46 [123 Cal.Rptr. 468, 539 P.2d 36].)
While our cases defining a bystander’s cause of action for negligent infliction of emotional distress consistently emphasize the first of the Row*687land factors—foreseeability of harm to plaintiff—discussion of the others has been limited. The second, fourth, and fifth factors may be disposed of quickly: certainty of injury is usually a jury question, particularly since we no longer require physical manifestations of mental distress (Hedlund v. Superior Court (1983) 34 Cal.3d 695, 706, fn. 8 [194 Cal.Rptr. 805, 669 P.2d 41, 41 A.L.R.4th 1063]); moral blame almost always militates in favor of recovery; and the policy of preventing future harm favors the plaintiff, but only slightly since, in most cases, any Dillon claim is simply added to the primary victim’s complaint. The third, sixth and seventh factors, however, merit more discussion.
This court has emphasized the importance of the third Rowland factor— nexus between defendant’s conduct and the risk of injury—in establishing limitations on recovery. In J’Aire Corp. v. Gregory (1979) 24 Cal.3d 799, 808 [157 Cal.Rptr. 407, 598 P.2d 60], we stated that case law “placefs] a limit on recovery by focusing judicial attention on the foreseeability of the injury and the nexus between the defendant’s conduct and the plaintiff’s injury.” There, we limited recovery for the tort of negligent interference with economic advantage “to instances where the risk of harm is foreseeable and is closely connected with the defendant’s conduct, where damages are not wholly speculative and the injury is not part of the plaintiff’s ordinary business risk.” (Ibid., italics added.)
The sixth and seventh Rowland factors—the burden on the defendant and the community, and the cost and availability of insurance—also merit further evaluation. Amici curiae2 contend that recovery in this case would mark an unwarranted expansion of Dillon, resulting in a new category of plaintiffs, fewer settlements, higher administrative costs and premiums, delays in payment, increased litigation, and higher awards. Amici insist that Dillon’s second guideline should be applied strictly, as a prerequisite for recovery. Apparently they have convinced the majority since the majority approvingly quote the reasoning in Borer v. American Airlines, Inc. (1977) 19 Cal.3d 441, 447 [138 Cal.Rptr. 302, 563 P.2d 858]: “We reasoned that we could not ‘ignore the social burden of providing damages . . . merely because the money to pay such awards comes initially from the “negligent” defendant or his insurer. Realistically the burden . . . must be borne by the public generally in increased insurance premiums or, otherwise, in the enhanced danger that accrues from the greater number of people who may choose to go without any insurance. We must also take into account the cost of administration of a system to determine and pay [the] awards; . . .’” (Maj. opn. at p. 665.)
*688The authorities upon which amici rely do not persuade me that Dillon has significantly contributed to any substantial increase in litigation and insurance premiums. Nor do I find any indication that other jurisdictions are retreating from Dillon (see 4 Speiser et al., The American Law of Torts (1987) §§ 16:25-16:26, pp. 1119-1126, and cases cited; Prosser & Keaton, Torts (5th ed. 1984) ch. 9, p. 366 et seq., and cases cited). As the Dillon court responded to the contention that otherwise meritorious claims should be barred out of fear of increases in the number of suits and of fraudulent claims: “ * “[We] should be sorry to adopt a rule which would bar all such claims on grounds of policy alone, and in order to prevent the possible success of unrighteous or groundless actions. Such a course involves the denial of redress in meritorious cases, and it necessarily implies a certain degree of distrust, which [we] do not share, in the capacity of legal tribunals to get at the truth in this class of claim.”’” (Dillon, supra, 68 Cal.2d at p. 744, quoting Hambrook v. Stokes Bros. (1925) 1 K.B. 141, quoting Dulieu v. White and Sons (1901) 2 K.B. 669, 681, opn. by Kennedy, J.; see also Molien v. Kaiser Foundation Hospitals (1980) 27 Cal.3d 916, 928-929 [167 Cal.Rptr. 831, 616 P.2d 813, 16 A.L.R.4th 518].)
I also do not believe courts lack the means to prevent unmeritorious cases from going to trial. As the case at bar demonstrates, trial courts are well aware of their duty to determine before trial whether the defendant could have owed the plaintiíf a duty of care under the facts. (See BAJI No. 12.84 (1986).)
III.
As for the instant case, I agree with Justice Benke of the Court of Appeal who in her concurring opinion indicated that the plaintiif mother’s claim should be permitted because “[u]nder current law, it cannot be said [plaintiff] is to be denied recovery because she did not see, hear or otherwise perceive the actual accident or injury. Case law permits the cause of action if she witnessed the immediate consequences of the event itself. It is a question of fact whether she did, and hence I would reverse the judgment below.” (Italics added.) Justice Benke declined to take a more liberal standard that “would extend liability far beyond existing limitations.” Her view is consistent both with Rowland, supra, 69 Cal.2d 108, and with the balance struck in Dillon, supra, 68 Cal.2d 728, between compensating legitimate claims and limiting liability.
I would conclude that the competing considerations cited in Rowland, (see ante, p. 686) as applied to this particular Dillon cause of action by the mother who sensorily perceived the immediate consequences of an accident involving her son, does not justify a departure from this state’s fundamental *689principle that a person is liable for injuries caused by his failure to exercise reasonable care under the circumstances. (Civ. Code, § 1714.) The majority’s strict requirement that plaintiff be present at the scene of the injury-producing event at the time it occurs and is then aware that it is causing injury to the victim will only bring about arbitrary results that will frustrate justice for victims of the negligent acts of others. We should apply the concepts of foreseeability and duty to negligent infliction of emotional distress actions, with a view toward a policy favoring reasonable restrictions on liability. This is a principled basis for determining liability and would also conform this area with other areas of negligence law.
The majority charge that, as former Supreme Court Justice Potter Stewart once said about obscenity, the Dillon guidelines mistakenly assumed that a court would know “duty” when it saw it. But more appropriate to the majority’s opinion and the arbitrary and unjust results it will soon engender is what Potter Stewart said, very simply, upon being appointed to the United States Supreme Court: “Fairness is what justice really is.” (The Young Justice (Oct. 20, 1958) Time, at p. 24.)
The majority declare “that a plaintiff may recover damages for emotional distress caused by observing the negligently inflicted injury of a third person if, but only if, said plaintiff: (1) is closely related to the injury victim; (2) is present at the scene of the injury producing event at the time it occurs and is then aware that it is causing injury to the victim; and (3) as a result suffers serious emotional distress—a reaction beyond that which would be anticipated in a disinterested witness and which is not an abnormal response to the circumstances.” (Maj. opn. at pp. 667-668, fns. omitted.)
Their brief was submitted by the National Association of Independent Insurers, the Association of California Insurance Companies, and the American Insurance Association.