I respectfully dissent.
Code of Civil Procedure section 340.6, subdivision (a) (section 340.6(a))1 provides that unless otherwise tolled, the statute of limitations for legal malpractice expires one year after the client’s actual or constructive discovery of the facts of the claim, or four years after the attorney’s wrongful act or omission, whichever occurs first. The majority conclude that for purposes of the one-year period, the defendant attorney bears the burden of proving when the plaintiff discovered, or should have discovered, the cause of action.
As the majority concede, however, this holding contravenes a long line of California decisions, including cases interpeting the identically structured *23limitations statute for medical malpractice (§ 340.5). These authorities consistently hold that where a limitations period runs from the time of the plaintiff’s discovery, he bears the burden of showing his suit was filed within the requisite time after discovery occurred. (E.g., Hobart v. Hobart Estate Co. (1945) 26 Cal.2d 412, 437 [159 P.2d 958] (Hobart) [fraud limitations statute]; April Enterprises, Inc. v. KTTV (1983) 147 Cal.App.3d 805, 833 [195 Cal.Rptr. 421] [breach of fiduciary duty]; Christ v. Lipsitz (1979) 99 Cal.App.3d 894, 898 [160 Cal.Rptr. 498] [current medical malpractice limitations statute]; Burgon v. Kaiser Foundation Hospitals (1979) 93 Cal.App.3d 813, 824 [155 Cal.Rptr. 763] [same]; Dujardin v. Ventura County Gen. Hosp. (1977) 69 Cal.App.3d 350, 355-356 [138 Cal.Rptr. 20] [same]; G. D. Searle & Co. v. Superior Court (1975) 49 Cal.App.3d 22, 25-26 [122 Cal.Rptr. 218] [product liability; latent injury]; DeVault v. Logan (1963) 223 Cal.App.2d 802, 809 [36 Cal.Rptr. 145] [predecessor medical malpractice limitations statute].)
Courts have given various rationales for the prevailing rule, but it is amply supported by the principle that one should usually not have to defend himself by proving facts peculiarly within his opponent’s knowledge. (See Cal. Law Revision Com. com., reprinted at 29B pt. 1 West's Ann. Evid. Code (1995 ed.) foll. § 500, p. 554.) This principle applies with particular force where a claim of legal malpractice is asserted. In such a case, the time of the plaintiff’s actual or constructive discovery may, and often will, depend on what information he obtained, and when he obtained it, from confidential and absolutely privileged consultations with another attorney. A party need not waive the privilege in order to help his opponent prove facts on which the adversary has the burden, and, for obvious reasons, there is little incentive to cooperate voluntarily. Indeed, as a result of the majority’s holding, it seems likely that in any retrial of this case, plaintiff Arlayna Samuels and Attorney Donald Hildre will invoke the attorney-client privilege to avoid testifying about the timing and content of their communications. Defendant Terence J. Mix may therefore be left without any opportunity to show Samuels’s claim is untimely.
I cannot join the vote to place an attorney sued for malpractice in such a legal and practical bind.
The issue here is whether the Legislature, when it adopted the legal malpractice limitations statute, intended an anomalous exception to the long-established rule that the plaintiff must demonstrate when his discovery of a cause of action triggered the running of the statute of limitations. The majority say such an exception arises on the face of the statute, appears consistent with legislative intent, and comports with public policy. I am not convinced.
*24Invoking the plain language of section 340.6(a), the majority distinguish the structure of this section from other statutory and common law rules of delayed discovery. The majority reason that in technical terms, section 340.6(a) defines the basic limitations period in terms of discovery, thus requiring the defendant to prove, as an element of his affirmative limitations defense, that the action is barred under the statutory terms. (Maj. opn., ante, at p. 7, citing Evid. Code, § 500.) By contrast, the majority argue, because belated discovery in other cases postpones the time at which the cause of action would “accrue” (see, e.g., § 338, subd. (d) [fraud or mistake]), the plaintiff may properly-have the burden of proving this “exception” to normal commencement and expiration of the limitations period. (Maj. opn., ante, at p. 14, quoting Hobart, supra, 26 Cal.2d 412, 437.)
The majority’s semantic analysis is overliteral and exaggerated. Section 340.6(a), like a “delayed accrual” rule, provides that even after malpractice and injury have occurred, the one-year limitations period will not begin until the plaintiff discovers his claim. Just as the defendant need not prove there was no delayed “accrual,” so he should not have to prove there was no delayed commencement.
The only practical distinction between section 340.6(a) and traditional delayed-discovery rules is that, with specified exceptions, section 340.6(a) imposes an absolute cutoff date of four years after malpractice and injury, regardless of discovery. But that particular feature of the statute has no logical relevance to the issue before us. Though the majority suggest otherwise, I see no reason why the statute’s provision of an alternate four-year limitations period should bear upon how to allocate the burden of proof of delayed discovery when the one-year period is at issue.
The majority assert there is no evidence of legislative purpose to adopt, for legal malpractice, the traditional allocation of proof of delayed discovery. But the majority have it backwards. The history of section 340.6(a), its language, and its subsequent construction by this court demonstrate that the Legislature intended no radical departure from traditional delayed-discovery rules. Indeed, the Legislature’s manifest aim was to adopt the common law delayed-discovery rule we had already announced for legal malpractice, subject only to exceptions we had invited the Legislature to impose.
The relevant background is set forth in the companion cases of Neel v. Magana, Olney, Levy, Cathcart & Gelfand (1971) 6 Cal.3d 176 [98 Cal.Rptr. 837, 491 P.2d 421] (Neel) and Budd v. Nixen (1971) 6 Cal.3d 195 [98 Cal.Rptr. 849, 491 P.2d 433] (Budd). At the time of those decisions, there was no distinct limitations period for attorney malpractice. Courts applied *25the two-year statute for “action[s] upon a contract, obligation or liability not founded upon an instrument of writing” (§ 339, subd. 1). (Neel, supra, at pp. 181-182.) The cause of action was deemed to accrue, and the limitations period thus began to run, when malpractice and resulting injury had occurred. (Neel, supra, at pp. 182-183; Budd, supra, at pp. 200-201.) In contrast with most other instances of professional negligence, however, courts declined to find that “accrual of the limitations period” was further delayed while the client lacked actual or constructive knowledge of the attorney’s negligence. (Neel, supra, at pp. 183-187.)
We found no justification for denying legal malpractice plaintiffs the benefit of the delayed-discovery rule. Thus, we held that a cause of action for legal malpractice did not accrue, and the limitations period thus did not begin to run (see § 312), until the client both discovered or should have discovered the malpractice, and suffered “actual and appreciable harm” thereby. (Neel, supra, 6 Cal.3d 176, 186-190; Budd, supra, 6 Cal.3d 195, 201.)
In Neel, we recognized that a rule of delayed accrual pending discovery could operate unfairly against legal malpractice defendants by extending the limitations period indefinitely. Neel thus expressly invited the Legislature to consider a solution already adopted for medical malpractice, i.e., an outside time limit that would apply regardless of the client’s failure to discover the professional negligence. (Neel, supra, 6 Cal.3d 176, 192-193 & fn. 32.)
The Legislature responded by enacting section 340.6 in 1977. Using a formula previously employed for medical malpractice (see § 340.5), section 340.6(a) established alternative limitations periods, one year from discovery, or four years from wrongdoing, whichever occurred first. Both of these limitations periods would be tolled while (1) the plaintiff had sustained no “actual injury” (id., subd. (a)(1)); (2) the negligent attorney continued to represent the plaintiff in the same legal matter (id., subd. (a)(2)); or (3) the plaintiff was under a legal or physical disability to sue (id., subd. (a)(4)). The four-year period, but not the one-year period, would further be tolled by the attorney’s willful concealment of the wrongful act or omission (id., subd. (a)(3); cf. Sanchez v. South Hoover Hospital (1976) 18 Cal.3d 93, 101 [132 Cal.Rptr. 657, 553 P.2d 1129] [medical malpractice]).
Thus, section 340.6(a) specifically retained the common law rules of Neel and Budd that commencement of the limitations period for legal malpractice occurs when a client has discovered, and has suffered injury from, an attorney’s professional malfeasance. In substance, section 340.6(a) departed from Neel and Budd in just two ways: It shortened the limitations period *26after discovery from two years to one; and, as Neel had suggested, it eliminated the unfair “long-tail” effect of the common law delayed-discovery rule by providing that failure to discover, unless caused by willful concealment, would not extend the statute of limitations past four years from the lawyer’s negligent act.
The majority point to no evidence, and I know of none, suggesting the Legislature had any broader goals in mind. We ourselves have recognized the limited aims, and the limited effects, of section 340.6(a). On several occasions, we have acknowledged that the 1977 statute essentially codified, with specified refinements, the “delayed accrual” rules of Neel and Budd. (Jordache Enterprises, Inc. v. Brobeck, Phleger & Harrison (1998) 18 Cal.4th 739, 749 [76 Cal.Rptr.2d 749, 958 P.2d 1062] [discovery and tolling provisions of section 340.6 invoke Neel and Budd; “[i]n section 340.6’s terms,” one-year limitations period that begins upon discovery is tolled pending actual injury]; Laird v. Blacker (1992) 2 Cal.4th 606, 611 [7 Cal.Rptr.2d 550, 828 P.2d 691] [section 340.6 “codified the discovery rule of Neel”]; see Cuadra v. Millan (1998) 17 Cal.4th 855, 865, fn. 11 [72 Cal.Rptr.2d 687, 952 P.2d 704] [listing section 340.6(a) among “rules postponing the accrual of a cause of action until a specified event occurs, e.g., until discovery of the facts” (italics in original)].)
But even if I agreed that the structure of section 340.6(a) is materially distinct from other common law and statutory delayed-discovery rules, I would apply the “escape clause” of Evidence Code section 500, and would thus retain, for attorney malpractice, the traditional burden of proving when the claim was discovered. As the Law Revision Commission’s comment to Evidence Code section 500 suggests, in deciding whether to depart from normal rules of burden allocation, “courts consider a number of factors: the knowledge of the parties concerning the particular fact, the availability of the evidence to the parties, the most desirable result in terms of public policy in the absence of proof of the particular fact, and the probability of the existence or nonexistence of the fact.” (Cal. Law Revision Com. com., reprinted at 29B pt. 1 West's Ann. Evid. Code, supra, foll. § 500, p. 554, italics added.)
The italicized portions of the passage quoted above apply directly here. As the majority appear to concede, the time the plaintiff actually learned of his or her claim, when different from the time the claim actually arose, is a matter primarily within the plaintiffs knowledge,2 The majority brush aside this concern with the comment that we do not “routinely” alter burdens of *27proof on this basis, and that plaintiffs are often forced to prove facts primarily within the defendant’s knowledge. (Maj. opn., ante, at p. 19.)
But there is merit in the notion that where a litigant has not willfully concealed information, his defense against a stale claim should not depend on his ability to prove when his opponent discovered that information, or should have done so. This principle supports the already prevailing rule that the burden of proving that commencement of a limitations period was delayed by failure to discover the claim is upon the person who seeks the benefit of the delay. Without compelling justification, the majority would depart, for purposes of attorney malpractice, from this established rule.3
In fact, if there is any case in which the plaintiff should assume the delayed-discovery burden, it is an action for legal malpractice. The defendant in such litigation faces unique and unfair difficulties if forced to prove the time of his opponent’s actual or constructive discovery. Here, in particular, the answer to the crucial questions—what plaintiff knew or suspected and when she knew or suspected it—is peculiarly and exclusively within plaintiff’s control.
This is because discovery of one lawyer’s malpractice will most often arise, as it did here, from the substance of the client’s consultations with another attorney. Proof of the time of discovery will thus depend, as it did here, on the content of those interviews. But such attorney-client communications are confidential and privileged by law. (Evid. Code, § 954.) Unless the client waives the privilege, neither he nor the attorney he consulted can be compelled to disclose the substance of their discussions.4
*28If the client bears the burden of proving when the malpractice claim was discovered, as the trial court ruled here, he may feel obliged, as plaintiff Samuels did here, to present evidence about the timing and nature of his consultations with a second lawyer. But if, as the majority hold, that burden rests with the attorney sued, there is no necessity, and no incentive, for the client to waive the privilege to aid his adversary in establishing a limitations defense. No lawyer worth his salt would allow his client to do so. Thus, it is unclear at best how an attorney sued for malpractice will be able to sustain his burden of proving when the client’s discussions with a second lawyer led to actual or constructive discovery of the malpractice claim.5
Accordingly, I disagree with the majority’s conclusion that the jury received an incorrect instruction about the burden of proof. But even if that instruction were erroneous, a conclusion also reached by the Court of Appeal, I am persuaded, contrary to the majority and the Court of Appeal, that the error was harmless on this record.
A malpractice statute of limitations triggered by actual or constructive discovery begins to run when the plaintiff knows she has suffered injury and suspects or has reason to suspect that professional blundering is its cause, whether or not she is aware of the precise legal theory or remedy by which redress is available. (E.g., Gutierrez, supra, 39 Cal.3d 892, 897-898 [medical malpractice].) Here, those requirements were clearly met more than one year before plaintiff Samuels filed this malpractice suit against defendant Mix.
*29The pertinent facts are essentially undisputed. Samuels suffers from eosinophilia myalgia syndrome, an incurable, progressively debilitating, and potentially fatal blood and muscle disorder. The disease has been linked to the drug L-tryptophan, which Samuels ingested. Samuels retained defendant Mix to pursue her remedies against Showa Denko, the drug’s manufacturer. Samuels hoped and believed the case was worth at least $2 million and was disappointed when Mix recommended she accept Showa Denko’s $400,000 settlement offer. Nonetheless, in 1990, Samuels did accept the offer, based on Mix’s advice that it was “very good,” and that Showa Denko, beset with other L-tryptophan cases, might enter bankruptcy.
The evidence, particularly including Samuels’s own deposition and trial testimony, further establishes the following: By October 1991, Samuels realized her $400,000 settlement with Showa Denko was insufficient, because the progress of her disease made clear she would never be able to work. She was also displeased with how Mix had handled the settlement, and had therefore lost confidence in him. Thus, in mid-October 1991, Samuels made an appointment with another attorney, Donald Hildre. Hildre, like Mix, specialized in L-tryptophan cases. Samuels met with Hildre on October 28, 1991. Her purpose was to determine whether she could “reopen” her case, and whether remedies were available against defendants other than Showa Denko. During the October 28, 1991, meeting, Hildre told her, contrary to Mix’s advice, that Showa Denko was financially sound and in no danger of bankruptcy. Samuels left the meeting believing that Mix had materially misrepresented Showa Denko’s financial status.6
In sum, no later than October 28, 1991, Samuels had concluded her settlement was inadequate, was actively exploring her remedies against all potential defendants, believed Mix had based his settlement advice on crucial misinformation, and distrusted Mix’s handling of her case for other reasons as well. By that date, Samuels was therefore on at least constructive notice that she had been the victim of professional negligence. Her suit, filed one year and two days later, was thus barred by the one-year provision of section 340.6(a). There is no reasonable probability the jury, if instructed *30that Mix rather than Samuels bore the burden of proving when discovery occurred, would have found otherwise. (People v. Watson (1956) 46 Cal.2d 818, 836 [299 P.2d 243]; see Soule v. General Motors Corp. (1994) 8 Cal.4th 548, 574 [34 Cal.Rptr.2d 607, 882 P.2d 298].) Any instructional error was therefore harmless.
For any or all of the reasons stated above, I would reverse the judgment of the Court of Appeal.
A11 subsequent statutory references are to the Code of Civil Procedure unless otherwise indicated.
The majority suggest there is no reason to assume constructive discovery is primarily within the plaintiff’s knowledge. Of course, the common law delayed-discovery rule, from *27which the majority here depart, does not make the distinction. In any event, I disagree with the majority’s premise. Constructive discovery of malpractice occurs as soon as the plaintiff suspects or has reason to suspect that professional negligence has occurred. (See Gutierrez v. Mofid (1985) 39 Cal.3d 892, 897-898 [218 Cal.Rptr. 313, 705 P.2d 886] (Gutierrez) [medical malpractice].) Thus, as with actual discovery, the analysis may often depend on events personal to the plaintiff, and his or her subjective reaction to those events. (Id., at p. 897.)
I do not accept the majority’s premise that defendants somehow deserve the burden of proving when attorney malpractice was discovered, because section 340.6(a) has benefitted them by eliminating the “long tail” of indefinite delayed-discovery rules. In any case where the statute’s one-year discovery period is involved, the four-year cutoff date is, by definition, not at issue. Moreover, section 340.6(a) already benefits legal malpractice plaintiffs by postponing commencement of the basic one-year limitations period while they excusably fail to discover their claims. Even if the alternate four-year period was intended to maintain fairness to defendants by eliminating the possibility of infinitely delayed claims, I cannot fathom why defendants should therefore be further penalized with respect to the one-year period.
Evidence Code section 958 provides a specific exception to the attorney-client privilege, under which a lawyer accused of malpractice may disclose his confidential communications with his disgruntled client insofar as they bear on the claim of professional negligence. But as *28we have noted many times, the general “litigant” exceptions to the physician-patient and psychotherapist-patient privileges (id., §§ 996, 1016) have no counterpart in the attorney-client privilege. Thus, the fact that communications with a lawyer may be probative on issues in a lawsuit, even issues injected by the client himself, does not preclude assertion of the privilege. (E.g., Brockway v. State Bar (1991) 53 Cal.3d 51, 63-64 [278 Cal.Rptr. 836, 806 P.2d 308]; People v. Caro (1988) 46 Cal.3d 1035, 1060, fn. 11 [251 Cal.Rptr. 757, 761 P.2d 680]; People v. Lines (1975) 13 Cal.3d 500, 509-516 [119 Cal.Rptr. 225, 531 P.2d 793].) And the specific exception allowing an accused attorney to defend himself by revealing his own confidential communications with the unhappy client does not expose confidential communications between the client and a different attorney later consulted about possible malpractice. (Schlumberger Limited v. Superior Court (1981) 115 Cal.App.3d 386, 392-393 [171 Cal.Rptr. 413].)
The majority suggest that privilege problems can be overcome by the simple device of asking the client, under oath, the date on which he learned the facts constituting his malpractice claim. The implication is that the client may not decline to answer such a question on grounds that the answer would reveal the substance of confidential communications with a second attorney. The assertion is not convincing. Even if the dates of a client’s meetings with his lawyer are not privileged, the client does have the privilege not to disclose what they discussed at any such meeting. If, as is likely, the nature and content of attorney-client discussions on a particular date are the means by which discovery occurred, the client may well be able to claim that by answering the question suggested by the majority, he would reveal the substance of confidential communications.
In Samuels’s deposition, read at trial, the following interchange occurred: “ ‘Question: After that meeting, that first meeting with Don Hildre, did you believe Terry Mix had done anything wrong in the handling of your case? ...[¶]... Answer: I don’t—you’re asking me if after the first time I met Don Hildre, did I think or did I believe that Terry Mix misrepresented to me? [¶] Question: Had done anything wrong in the handling of your case. [¶] Answer: Yes.’ ” At trial, Samuels claimed that if, before meeting with Hildre, she already felt Mix had done something “wrong,” her dissatisfaction arose solely from two peripheral complaints, i.e., he used a courier to deliver a settlement check, and he discussed the settlement with others, though Samuels understood it was confidential. Both at trial and in her deposition, Samuels further explained that as a result of the October 28, 1991, meeting with Hildre, she believed Mix had “misrepresented” the financial strength of Showa Denko.