I respectfully dissent. As will appear, the standardized “personality function” tests at issue here, if used in the manner proposed by the two defendants in this case, are in all essential respects indistinguishable from polygraph, voiceprint or other scientific tests aimed at proving, or disproving, guilt. For that reason, these standardized tests must meet the usual Kelly/Frye standard for admissibility.
I readily acknowledge the admissibility of opinion evidence tending to show that the defendant’s character or trait of character makes it unlikely that he or she committed a charged offense. (See Evid. Code, § 1102, subd. (a); People v. Jones (1954) 42 Cal.2d 219, 222 [266 P.2d 38] [defendant lacked usual characteristics of sex deviants].) I also recognize that such evidence may be presented in the form of expert testimony (see Evid. Code, § 801), which ordinarily is not subject to the Kelly/Frye admissibility standards (see People v. McDonald (1984) 37 Cal.3d 351, 372-373 [208 Cal.Rptr. 236, 690 P.2d 709, 46 A.L.R.4th 1011]).
But the present case involves more than a simple expert opinion to the effect that defendant Grafton was a “normal” heterosexual whose character lacked the potential for sexual deviation. Instead, Dr. Mitchell, defendant’s expert, was prepared to testify that he largely based his opinion of normality on certain standardized written tests, self-administered to defendant in her jail cell, which supposedly measured the subject’s “general possibilities for abnormality in psychological function.” According to Dr. Mitchell, the primary test, the Minnesota Multiphasic Personality Inventory (MMPI), consisting of 566 written questions, had several built-in “validity scales” designed to assure accuracy and detect lies by the person taking the exam. Dr. Mitchell observed that many experts believe the test makes it virtually impossible to conceal an abnormal personality profile.
After reviewing the test results and interviewing defendant in her cell for two hours, Dr. Mitchell was prepared to state his opinion that she had a “normal personality function,” and “is falsely charged in this matter.” He *1164qualified his latter response by opining that defendant probably has not “engaged in the past in sexual deviancy of any kind,” and “it is unlikely . . . she would be involved in” the charged offenses. Dr. Mitchell testified that the scientific community generally accepts the MMPI as a valid measure of psychopathology. But, as the majority explains, the prosecutor’s attempts to learn the scope of Dr. Mitchell’s knowledge in this regard were “inconclusive.” (Maj. opn., ante, at p. 1150.)
On cross-examination, the prosecutor developed the following: Dr. Mitchell could identify no studies indicating whether child abusers have ever registered “normal” on the subject tests. He conceded that some experts believe the MMPI may not identify “situational” child abusers, i.e., those offenders reacting to situational stress, rather than chronic, long-term abusers. Indeed, Dr. Mitchell acknowledged that persons acting under extreme embtional stress may not score reliably on that test. He also confirmed ¡that although the reliability of these tests depends on the circumstances under which they are administered, neither he nor anyone else was present when the tests were completed by defendant Grafton in her jail cell.
Dr. Mitchell was unable to cite any studies on the validity of the subject tests as applied to individuals charged with either child abuse or incest. He acknowledged that on one occasion an admitted child molester had registered “normal” on an MMPI test he had given.
On this record, we might reasonably hold that the trial court, without abusing its discretion, could have excluded the evidence as unduly prejudicial to the People under Evidence Code section 352, or that in any event no prejudice to defendant could have resulted from its exclusion. I confine my dissenting remarks, however, to the Kelly/Frye issue. In my view, the concessions and omissions disclosed through cross-examination of Dr. Mitchell lend substantial force to the trial court’s ruling that the subject tests were inadmissible because they failed to meet the Kelly/Frye standard of general acceptance (in the scientific community.
The trial court excluded the proffered testimony on Kelly/Frye grounds because the defense failed to show the standardized test on which Dr. Mitchell relied was generally accepted in the scientific community. I would affirm that ruling.
I note initially that expert testimony is subject to the overriding requirement of reasonable reliability. Under present law, expert opinion evidence is limited to opinions based on matter perceived by or personally known to the witness “that is of a type that reasonably may be relied upon by an expert in forming an opinion . . . .” (Evid. Code, § 801, subd. (b).) The court is *1165entitled to examine the witness, prior to testifying, concerning the matter on which his opinion is based (id., § 802), and to exclude opinion testimony based on matter “that is not a proper basis for such an opinion” (id., § 803).
In my view, Kelly/Frye is unaffected by Proposition 8 (adding article I, section 28, subdivision (d) to the state Constitution). I doubt this “Truth-in-Evidence” provision was intended to deprive the trial courts of their traditional power to exclude unreliable scientific or expert testimony. By its terms, the provision is limited to “relevant” evidence, and evidence failing to meet Kelly/Frye's reliability standard reasonably could be deemed irrelevant to the issues. I also note that the provision recites that it was not intended to affect any existing statutory rule of evidence relating to hearsay, or Evidence Code sections 352 or 1103. As the majority acknowledges, these exceptions may indicate an intent to preserve the general character or opinion evidence restrictions. (See People v. Harris (1989) 47 Cal.3d 1047, 1082 [255 Cal.Rptr. 352, 767 P.2d 619].)
The Kelly/Frye test is merely an application of the “reasonable reliability” standard to purported “scientific” methods of establishing guilt or innocence. Contrary to the majority’s implication, nothing in our opinions in People v. McDonald, supra, 37 Cal.3d 351, or People v. Kelly (1976) 17 Cal.3d 24 [130 Cal.Rptr. 144, 549 P.2d 1240], would flatly exempt expert testimony of the kind involved here, where such testimony rests in large part on the results or readings of a scientific test, process or procedure.
In Kelly, involving voiceprint identification procedures, we observed that generally, admissibility of expert testimony based on a new scientific technique involves a two-step process whereby (1) the reliability of the expert witness’s methods are established, and (2) the witness is properly qualified as an expert to state an opinion on the subject. (17 Cal.3d at p. 30.) Kelly adopted the test for reliability of the technique itself from Frye v. United States (D.C. Cir. 1923) 293 Fed. 1013, 1014 [54 App.D.C. 46, 34 A.L.R. 145]: The technique “must be sufficiently established to have gained general acceptance in the particular field in which it belongs.”
Kelly noted that the foregoing “general acceptance” standard is particularly appropriate when the scientific evidence, presented by experts, is apt to be given considerable, perhaps undue, weight by lay jurors impressed with the “posture of mystic infallibility” which such evidence frequently may present. (17 Cal.3d at pp. 31-32, quoting an earlier case.) Kelly held that neither case law nor scientific articles satisfactorily established that the voiceprint technique had been generally accepted in the scientific community.
*1166In McDonald, we held simply that the trial court erred in excluding defense expert testimony on the psychological factors that might undermine the accuracy of eyewitness identifications. We declined to apply Kelly/Frye to such expert testimony, noting that this standard has never been applied to expert testimony of that kind. Significantly, in McDonald we distinguished between general expert opinion testimony (where the jurors can employ “healthy skepticism” to temper their acceptance of the proffered opinion) and evidence presented by an expert but produced by a “machine,” i.e., “an apparently ‘scientific’ mechanism, instrument or procedure'" (where the “aura of infallibility that often surrounds such evidence may well conceal the fact that it remains experimental and tentative”). (37 Cal. 3d at pp. 372-373, italics added.) Of course, as we made clear in People v. Shirley (1982) 31 Cal.3d 18 [181 Cal.Rptr. 243, 723 P.2d 1354], Kelly/Frye is not limited to “machines” or techniques that involve the manipulation of physical evidence: Testimony based on a new scientific process “operating on purely psychological evidence” would be subject to the Kelly/Frye test. (Id. at p. 53.)
Close on point is People v. Bledsoe (1984) 36 Cal.3d 236 [203 Cal.Rptr. 450, 681 F.2d 291], in which we found error in admitting evidence of “rape trauma syndrome” to assist in proving that a rape occurred. A careful reading of that case reveals, contrary to the majority’s interpretation herein, a strong implication that diagnosis of the syndrome constituted a new scientific technique which must satisfy the Kelly/Frye test. As we held in Bledsoe, “Because the literature does not even purport to claim that the syndrome is a scientifically reliable means of proving that a rape occurred, we conclude that it may not properly be used for that purpose in a criminal trial.” (Id.j at p. 251, italics added.) The diagnostic tests at issue here are essentially indistinguishable. Indeed, in my view, the need for Kelly/Frye compliance is even more compelling than in Bledsoe, for unlike rape trauma syndrome evidence, the “personality function” test results proffered here were offered not merely to prove that a crime occurred, but to prove that defendant Grafton could not have committed it. In other words, the essential thrust of the “character” evidence was directed to the crucial issues of guilt or innocence.
The majority nonetheless disputes the application of Kelly/Frye to the standardized tests that formed the basis for Dr. Mitchell’s testimony. The majority observes that “[t]he methods employed are not new to psychology or the law, and they carry no misleading aura of scientific infallibility.” (Ante, at p. 1157.) I disagree on both counts.
First, in my view a new application of these standardized tests is indeed involved here. As indicated previously, I acknowledge that California *1167courts have allowed expert opinion on the subject of a defendant’s mental state, including character evidence indicating the likelihood that he could not have committed the charged offenses. Moreover, I realize that some courts have admitted expert opinion based on standardized tests, similar to those administered here, to form a psychiatric diagnosis of a defendant’s mental state at the time the alleged acts were committed. (See maj. opn., ante, at p. 1154, and cases cited.) But using such tests to formulate a psychiatric diagnosis of the defendant’s mental state is far different than using them to exclude defendant from the relevant class of offenders in much the same manner as a blood test or voiceprint.
Dr. Mitchell indicated that these tests are used in clinical practice as diagnostic and treatment tools, but he did not suggest that they are commonly used in criminal proceedings to identify, or exclude, deviant personalities. As stated in the majority opinion of the Court of Appeal herein, “Dr. Mitchell. . . failed to establish the general acceptance of reliability [of the testing at issue] in the particular field to which it belongs. A test may be widely utilized as a tool or a guide while recognizing and accepting the fact the procedure lacks the desired certainty.”
As conceded by defense counsel at oral argument, no cases exist which sanction such a revolutionary application of standardized test procedures as was attempted in this case. I think it ironic that all three of the cases from our court cited by the majority herein as supporting reliance on standardized psychological tests, such as the MMPI involved here, contained references to doubts or concerns regarding the value of such tests. In People v. Coogler (1969) 71 Cal.2d 153, 166-167 [77 Cal.Rptr. 790, 454 P.2d 686], in which the prosecution produced no evidence to rebut the defendant’s showing of diminished capacity (based in part on the MMPI), we observed that “a jury could properly reject the [defense] expert’s conclusions because of doubt as to the material upon which these conclusions were based.” In People v. Cruz (1980) 26 Cal.3d 233, 248 [162 Cal.Rptr. 1, 605 P.2d 830], we noted that one of the psychiatrists testifying at the defendant’s trial “explained to the jury why he did not use [standardized] psychological tests: He believed, based on years of experience, that they were not reliable. ” (Italics added.) And in People v. Coleman (1985) 38 Cal.3d 69, 80 [211 Cal.Rptr. 102, 695 P.2d 189], we pointed out that a testifying psychiatrist “characterized the use of psychological tests such as the . . . Minnesota Multiphasic Personality Inventory as lacking in reliability and inappropriate for use in forming an opinion on whether an accused possessed a certain mental state at some time in the past.” (Italics added; see also People v. John W. (1986) 185 Cal.App.3d 801, 804-805 [229 Cal.Rptr. 783] [expert testimony doubting utility of MMPI in detecting deviant sexuality].)
*1168Contrary to the majority’s implication herein, none of the foregoing cases indicated that the MMPI or similar standardized psychological tests were generally accepted by the scientific community, and none involved attempts to prove the defendant’s factual innocence through the guise of expert “character” evidence. Indeed, the only case I have found in which such an attempt was made resulted in an affirmance of a ruling excluding the evidence. (People v. John W., supra, 185 Cal.App.3d at p. 806.) Thus, I would conclude the standardized tests relied on by Dr. Mitchell do indeed involve a “new,” Unprecedented scientific technique within the proper scope of Kelly/Frye.,
Second, as for Kelly/Frye's element of an “aura of infallibility” surrounding the disputed technique, Dr. Mitchell’s offer of proof included his opinion that thq tests he relied on were accurate, having a better than 70 percent chance of being right and, according to some experts, being virtually impossible to conceal one’s deviant character from the tester. Dr. Mitchell described as á “fair characterization” a colleague’s observation that “achieving a normal profile in the MMPI is like walking through a minefield and coming out intact.” Significantly, according to Dr. Mitchell, built-in “validity scales” supposedly assured that no intentional concealment of a deviant personality i could occur. Quite clearly, a jury could be unduly swayed by such claims of accuracy and invincibility.
It is appdrent that, based on Dr. Mitchell’s description, the standardized tests combined elements of (1) a lie detector, aimed at uncovering untruthful test responses, and (2) a “personality print,” i.e., a supposedly foolproof scientific method of matching a suspect’s personality traits with the offense in question; much in the same way that voiceprints, fingerprints, blood or semen samples, are used to match or exclude a suspect.
The majority concedes that polygraph and voiceprint techniques are unquestionably subject to Kelly/Frye. In my view, Dr. Mitchell’s “validity scales” and “personality print” should be subject to the same scrutiny. (See also Seering v. Department of Social Services (1987) 194 Cal.App.3d 298, 311 [239 Cal.Rptr. 422] [child sexual abuse accommodation syndrome subject to Kelly/Frye]; In re Amber B. (1987) 191 Cal.App.3d 682, 690-691 [236 Cal.Rptr. 623] [use of anatomically correct dolls to support psychiatric opinion of child molestation subject to Kelly/Frye].) In both Seering and Amber B. the courts stressed the likelihood that a jury would place undue weight on the scientific nature of the underlying evidence.
I do not suggest the results of these tests are necessarily inadmissible. The purpose of Kelly/Frye is not to exclude evidence but to test its reliability by giving the members of the scientific community “the determinative voice” *1169in deciding that issue. (See People v. Kelly, supra, 17 Cal.3d at p. 31, quoting from United States v. Addison (D.C. Cir. 1974) 498 F.2d 741, 743-744 [162 App.D.C. 199].) There may well be a body of scientific research showing that the standardized measures of “personality function” are generally accepted by mental health experts as scientifically reliable tests of sexual deviancy or normality. On this record, however, such evidence is conspicuously absent.
With due respect, I am concerned about the potential consequences arising from the majority’s holding. I foresee complicated and time consuming “mini-trials” devoted not to the factual issue of guilt or innocence of the accused, but focusing instead on his “personality profile” and its correlation with the profile displayed by the average child molester, robber, arsonist, or whomever. Presumably, the People would have the opportunity to rebut the defendant’s showing, using their own experts and standardized tests. (See Evid. Code, § 1102, subd. (b).) The prospects for sidetracking the trial’s “main event” are enormous.
I would affirm the Court of Appeal’s decision upholding defendant’s conviction.
Kaufman, J., concurred.