We granted review in this case to resolve a conflict in the Courts of Appeal concerning the admissibility of battered women’s syndrome (BWS) testimony pursuant to Evidence Code section 11071 in the absence of evidence of a prior incident of domestic violence. *909(Compare People v. Williams (2000) 78 Cal.App,4th 1118 [93 Cal.Rptr.2d 356], with People v. Gomez (1999) 72 Cal.App.4th 405 [85 Cal.Rptr.2d 101].) Rather than confront this question directly, the majority holds that the testimony in this case was, in any event, admissible under the more general provisions of section 801 concerning expert witnesses. In reaching this conclusion, the majority not only evades determining the proper application of section 1107 and our responsibility to “secure uniformity of decision” (Cal. Rules of Court, rule 28(b)(1)), but disregards settled principles governing expert testimony. Because I find the evidence regarding BWS was not relevant to any disputed fact—and its admission prejudiced defendant—I dissent.
I
In determining the admissibility of any expert witness testimony, the threshold question is not simply whether the subject “is sufficiently beyond common experience that the opinion of an expert would assist the trier of fact . . . .” (§ 801, subd. (a).) Rather, as with all evidence, the first inquiry is whether the testimony is relevant, i.e., whether it has “any tendency in reason to prove or disprove any disputed fact that is of consequence to the determination of the action,” including the credibility of a witness. (§ 210.) Unless and until the proponent demonstrates that an expert’s testimony has the requisite relationship to a disputed fact, it is irrelevant and cannot assist the trier of fact.
“Battered women’s syndrome ‘has been defined as “a series of common characteristics that appear in women who are abused physically and psychologically over an extended period of time by the dominant male figure in their lives.” [Citations.]’ [Citation.]” (People v. Humphrey (1996) 13 Cal.4th 1073, 1083-1084 [56 Cal.Rptr.2d 142, 921 P.2d 1] (Humphrey); People v. Erickson (1997) 57 Cal.App.4th 1391, 1399 [67 Cal.Rptr.2d 740].) Accordingly, the overwhelming majority of courts considering the question have concluded that the relevance of expert testimony regarding BWS depends upon its proponent establishing “some evidentiary foundation that a party or witness to the case is a battered woman, and that party or witness has behaved in such a manner that the jury would be aided by expert testimony providing an explanation for the behavior. [Citation.]” (State v. Borrelli (1993) 227 Conn. 153, 172, fn. 15 [629 A.2d 1105, 1115]; see Fennell v. Goolsby (E.D.Pa. 1985) 630 F.Supp. 451, 459; State v. Niemeyer (1999) 55 Conn.App. 447, 452 [740 A.2d 416, 419-420]; State v. Stringer (1995) 271 Mont. 367, 378 [897 P.2d 1063, 1070]; State v. Koss (1990) 49 Ohio St.3d 213, 218 [551 N.E.2d 970, 974-975]; Bechtel v. State (Okla.Crim.Ct.App. 1992) 1992 OKCR 55 [840 P.2d 1, 9]; State v. Ogden (2000) 168 Or.App. 249, 257 [6 P.3d 1110, 1114]; see also State v. Kelly (1984) 97 N.J. 178, 200-201 [478 A.2d 364, 375].)
*910Here, the record contained no evidence Kimberly Pipes suffered from BWS. Not only had there been no pattern of physical or psychological abuse in her relationship with defendant, she flatly denied he had ever hit her in the past. Nor were there any of the substantive characteristics—dependency, control, isolation—that mark BWS. Nevertheless, and contrary to the majority’s reading of the record (see maj. opn., ante, at p. 901), expert witness Jeri Darr testified at length as to general causes and effects of BWS as well as anecdotal BWS experiences she had observed in the course of her work with domestic violence victims.2 She described in considerable detail the cycle of tension-building, explosion of violence, and contrition that typifies the battering relationship, and the factors that contribute to its perpetuation, such as the victim’s dependence on the abuser; the abuser’s desire for power and control, his isolation of the victim, and his threats of harm to the victim and/or other family members; and the victim’s passivity, fear, and lack of self-esteem. Darr also explained the common reactions of women in these abusive relationships; denial or minimizing of physical injury, self-blame, reconciliation and reunion with the abuser, learned helplessness, and recantation.
As the majority acknowledges, “[t]he prosecutor did not claim that defendant had beaten the victim before or that defendant fit the profile of a batterer” (maj. opn., ante, at p. 902), or that Pipes suffered BWS. Thus, general BWS evidence was plainly irrelevant to any disputed issues. (See, e.g., State v. Ogden, supra, 168 Or.App. at p. 257 [6 P.3d at p. 1114].) To paraphrase the court in Ogden, “at trial, the state failed to establish a critical link between the expert’s testimony about why a battered woman might [recant her earlier claim of abuse] and why this complainant did so. Specifically, the state did not establish that complainant suffers from BWS. If complainant does not suffer from BWS, then testimony about that subject seemingly has no bearing *911on the complainant’s behavior.” (Id. at p. 256 [6 P.3d at p. 1114].) Thus, in the absence of evidence Pipes exhibited BWS, a substantial portion of Darr’s testimony was irrelevant and should not have been admitted whether the evidentiary basis was section 1107 or section 801.
II
The majority concludes expert evidence was nevertheless relevant to counter certain misconceptions lay jurors may have about the behavior of victims of generic domestic violence. At the hearing on defendant’s motion in limine, the prosecutor indicated she intended to introduce Darr’s testimony “to explain why [Pipes] recanted at the preliminary hearing, and I anticipate that she will recant at the trial and why she gave the inconsistent statements and why, frankly, she went back to him.” Even if the testimony had been restricted to the recantation and reconciliation aspects of violent domestic situations,3 I would still take issue with its relevance. The majority’s analysis is equally untenable even on this more limited question.
According to some experts, including Darr, even first-time victims of domestic violence are likely in certain circumstances to recant their initial claims of abuse or threats of harm. From this premise, the argument goes that since Pipes recanted some of her original statements to the police and since she was a victim of domestic violence, her recantation was likely false. The syllogism is faulty in several respects. It assumes that generic “domestic violence”—like BWS—describes a pattern of behavior that includes certain reactions to incidents of abuse. In reality, “domestic violence” is neither cause nor effect; it is simply a label, now codified (see Fam. Code, § 6211),4 for abuse occurring in statutorily specified relationships. Moreover, the argument *912essentially works backward from effect to posit cause: That is, even in the absence of any history of abuse between defendant and Pipes, the fact of their domestic relationship coupled with her denial at trial of any assault or threat of harm made the substance of her recantation more likely false. Without, however, some evidence Pipes came within Darr’s “domestic violence victim” construct, testimony as to the significance of Pipes’s recantation had no evidentiary foundation. (See State v. Ogden, supra, 168 Or.App. at p. 256 [6 P.3d at p. 1114].) It does not, as the prosecutor asserted it would, explain “why.” Recantation is not a predicate of abuse but the other way around. Only when a woman is in a relationship exhibiting characteristics of domestic violence can expert evidence elucidate its effects on her behavior and the reasons she may be more likely to recant an initial claim of assault.
In sum, as defense counsel explained at oral argument, the problem with the majority’s reasoning is that under section 801, admissibility depends upon having some evidence in the record consistent with the subject matter of the expert’s testimony other than the underlying incident. (Cf. Humphrey, supra, 13 Cal.4th at pp. 1096, 1098-1100 (cone. opn. of Brown, J.).) In this context, the foundational facts must consist of the substance, not the manifestation, of an abusive domestic relationship as described by the expert, and not simply as a matter of statutory definition. No such evidence appears here.5 Although defendant and Pipes had known each other and dated off and on for 11 years and had been living together intermittently at the time of the incident, defendant had never abused or threatened Pipes or her children. Nor does the record establish that Pipes was in any respect dependent on defendant emotionally or financially, that he exerted control over her, or that he kept her isolated from family and friends. Pipes does not appear to be a passive personality and was unequivocal at trial that she reacted out of anger, and not fear, when she initially accused defendant.
*913On these facts, nothing in the majority’s analysis distinguishes a first-time or single-incident assault committed in a domestic relationship from one committed in a nondomestic relationship. Recantation is hardly unique to domestic situations (see, e.g., People v. Sam (1969) 71 Cal.2d 194 [77 Cal.Rptr. 804, 454 P.2d 700]; People v. Anthony O. (1992) 5 Cal.App.4th 428 [6 Cal.Rptr.2d 794]; People v. Plasencia (1985) 168 Cal.App.3d 546 [223 Cal.Rptr. 786]; see also People v. Odom (1969) 71 Cal.2d 709 [78 Cal.Rptr. 873, 456 P.2d 145]; People v. Loyd (1977) 71 Cal.App.3d Supp. 1 [139 Cal.Rptr. 693]); and the majority fails to identify what misconceptions lay jurors may harbor in these circumstances that require expert testimony to correct. (Cf. People v. Bothuel (1988) 205 Cal.App.3d 581, 587 [252 Cal.Rptr. 596], disapproved on other grounds in People v. Scott (1994) 9 Cal.4th 331, 348 [36 Cal.Rptr.2d 627, 885 P.2d 1040].)
The attempt to analogize to expert testimony concerning rape trauma syndrome (RTS) or child sexual abuse accommodation syndrome (CSAAS) is unpersuasive. (See People v. Bledsoe (1984) 36 Cal.3d 236, 247-248 [203 Cal.Rptr. 450, 681 P.2d 291]; People v. Sanchez (1989) 208 Cal.App.3d 721, 735 [256 Cal.Rptr. 446]; cf. People v. McAlpin (1991) 53 Cal.3d 1289, 1300-1302 [283 Cal.Rptr. 382, 812 P.2d 563].) The principal distinction is that misconceptions about the reactions of rape or child molestation victims are much more likely to arise from an initial or single incident; and neither rape nor molestation depends upon the nature of the relationship between the defendant and the victim or particular characteristics of that relationship. The same cannot be said regarding an initial incident of generic domestic violence. “Any woman may find herself in an abusive relationship with a man once.” (Walker, The Battered Woman Syndrome (1979) p. XV; cf. People v. Gomez, supra, 72 Cal.App.4th at pp. 416-417.) Misconceptions arise when the woman seemingly allows the relationship to continue despite repeated abuse, since uninformed jurors may assume she would leave when the opportunity presented itself. (See, e.g., Humphrey, supra, 13 Cal.4th at p. 1087; State v. Kelly, supra, 478 A.2d at p. 370.) In other words, in situations of domestic violence, it is the repetitive aspect of the abusive behavior that becomes relevant in terms of the victim’s counterintuitive state of mind and reactions. With RTS and CSAAS, the initial sexual assault or molestation is sufficient to trigger the reactions—such as delayed reporting— about which the expert enlightens the jury.
Moreover, in McAlpin, Bledsoe, and other cases involving rape and child molestation, the expert testimony was in virtually all cases admitted to explain counterintuitive out-of-court conduct, not to attack the credibility of a victim’s sworn trial testimony. (Cf. People v. Morgan (1997) 58 Cal.App.4th 1210, 1216 [68 Cal.Rptr.2d 772]; but see People v. Housley (1992)
*9146 Cal.App.4th 947, 955-956 [8 Cal.Rptr.2d 431].) In the former situation, no rules of evidence are available to rebut misconceptions without expert testimony. In the latter, section 780 and, in particular, section 1235 provide evidentiary bases both to challenge the victim’s credibility and to proffer her out-of-court statements as substantive evidence.
in
Contrary to the majority’s implication, excluding expert testimony on these facts does not somehow allow a batterer to act with impunity. (Maj. opn., ante, at p. 904, quoting People v. Williams, supra, 78 Cal.App.4th at p. 1129.) It simply means the prosecutor may not use expert testimony to attack the credibility of the victim’s sworn testimony or to bolster the credibility of her hearsay reports to law enforcement or others. Nonetheless, her prior inconsistent statements are admissible as substantive evidence (see § 1235; People v. Chavez (1980) 26 Cal.3d 334, 354-361 [161 Cal.Rptr. 762, 605 P.2d 401]); and she may be impeached on a wide variety of bases, including her demeanor; bias, interest, or other motive; prior inconsistent statements; and attitude toward the proceedings (see generally § 780, subds. (a), (f), (h), (j)). The majority offers no rationale for concluding the prosecutor should not be restricted to these standard rules of evidence in single-incident domestic assault cases just the same as in nondomestic assault cases. Clearly, in the latter circumstance, the prosecution would not be entitled to invoke expert testimony to “explain” why the victim’s testimony under oath should be discounted in favor of her statement to the police at the time of the alleged act. Furthermore, at least with respect to an initial or single assault, nothing in this record establishes that lay jurors are any less able in the absence of expert testimony to understand why a victim might falsely recant in the former as in the latter. That being the case, admission of expert testimony runs a substantial risk of invading the exclusive factfinding province of the jury (cf. Humphrey, supra, 13 Cal.4th at pp. 1099-1100 (cone. opn. of Brown, J.)) and undermining the trial court’s gatekeeper function in exercising discretion to exclude such evidence for that reason. (§ 352; see, e.g., People v. Roscoe (1985) 168 Cal.App.3d 1093, 1100 [215 Cal.Rptr. 45].)
The majority’s holding creates a special rule for the admission of expert testimony when the defendant and alleged victim of an assault have or have had a personal relationship that comes within the definition of “domestic violence” set forth in Family Code section 6211—a “domestic situation” exception to the usual relevancy requirement.
*915Among the problems with such an exception, it effectively evades application of section 1107 and, in particular, the express limitation on the use of BWS evidence “against a criminal defendant to prove the occurrence of the act or acts of abuse which form the basis of the criminal charge.” (§ 1107, subd. (a); cf. People v. Bledsoe, supra, 36 Cal.3d at p. 251 [RTS evidence not admissible to prove victim was raped].) As defense counsel explained at oral argument, the expert’s testimony on the question of recantation is relevant only if Pipes were a victim of domestic violence so that her experience was consistent with Darr’s expertise regarding such circumstances. But the only incident of domestic violence was the underlying charge. Therefore, the jury must assume the truth of the criminal allegation to accept Darr’s testimony as to the significance of Pipes’s recantation. Absent enforcement of the statutory restriction, section 1107 is rendered nugatory.
The majority’s holding may also become an unwieldy double-edged sword that invites a host of unintended consequences. Section 1107 permits introduction of expert testimony regarding BWS for both the prosecution and defense. Nothing in the majority’s reasoning suggests any basis for a different standard of admissibility for defense expert testimony under section 801 or limits the defense from similarly introducing non-BWS domestic violence evidence based on only a single battering incident. Nor would such evidence be restricted to testimony about recantation or reconciliation as in this case. Moreover, Family Code section 6211 broadly defines “ ‘[d]omestic violence’ ” to include a wide range of individuals and relationships. The majority’s same reasoning may well be asserted in seeking to introduce new strains of expert evidence with respect to these as well as the more common situation at issue here. Such a result would bring expert testimony far afield from the BWS evidence admissible under section 1107, particularly given the consensus that the counterintuitive behavior of battered victims follows from the repeated cycle of violence they learn to accommodate. (See Humphrey, supra, 13 Cal.4th 1073, 1096 (cone. opn. of Brown, J.); People v. Gomez, supra, 72 Cal.App.4th at pp. 416-417; State v. Ogden, supra, 168 Or.App. at p. 257 [6 P.3d at p. 1114]; Ibn-Tamas v. United States (D.C.Ct.App. 1979) 407 A.2d 626, 634.)
IV
Since I conclude the trial court should have excluded Darr’s testimony in its entirety, the question remains whether its admission resulted in a miscarriage of justice. (People v. Watson (1956) 46 Cal.2d. 818, 835 [299 P.2d 243].)
*916Although this is a close case, I would find prejudice for the following combination of reasons: The charge of making terrorist threats against Pipes (Pen. Code, § 422)—and arguably the false imprisonment charge involving her (id., § 236)—rested solely on Pipes’s statements to law enforcement. Dan-testified extensively regarding BWS, well beyond the limited purpose of explaining recantation or reconciliation. In her closing argument, the prosecutor similarly did not restrict herself to the proffered justification, but heavily adverted to the whole of Dan’s testimony. The trial court likewise did not admonish the jury to confine its consideration of the evidence to these factors. The jury failed to return verdicts on similar charges involving Canie Miller, who also recanted, but whose out-of-court statements did not have the benefit of any BWS or domestic violence credibility enhancement. Finally, courts have long recognized the powerful impact of expert testimony in jury deliberations, for the very reason that such evidence relates to matters outside the layperson’s ken. (See, e.g., People v. Gomez, supra, 72 Cal.App.4th at p. 418; see also People v. Axell (1991) 235 Cal.App.3d 836, 862 [1 Cal.Rptr.2d 411].)
In light of the foregoing, I would reverse the judgment at least as to the violation of Penal Code section 422.
All undesignated statutory references are to the Evidence Code.
Contrary to the majority’s statement that “neither Darr’s testimony nor the prosecutor’s argument [to the jury] was premised on the admissibility of her testimony under section 1107” (maj. opn., ante, at p. 902), the record is replete with references to “battered women’s syndrome” and by implication section 1107. At the hearing on the motion in limine to exclude Darr’s evidence, the court referred to “testimony about battered women’s syndrome”; and the prosecutor and defense counsel argued its admissibility or inadmissibility, respectively, under People v. Williams, supra, 78 Cal.App.4th 1118 and People v. Gomez, supra, 72 Cal.App.4th 405, both premised on section 1107. At the close of the hearing, the prosecutor specifically stated, “As Williams indicates, there’s no limitation in . . . Evidence Code [section] 1107, that there be prior incidents, and that’s why the Williams case allowed it in.” In discussing the evidence during closing argument, the prosecutor asked the jury, “How did we know it [the incident] happened? . . . [f] First of all, the victim made the statement to the deputy shortly after the incident. . . . And we also heard from Jeri Darr, domestic violence expert who spoke about the battered women’s syndrome.” The trial court twice instructed the jury on the matter: immediately prior to Darr’s testimony—“Evidence may be presented to you concerning battered women’s syndrome”—and at the conclusion of the evidence—“Evidence has been presented to you concerning battered women’s syndrome.” Finally, defendant appealed his conviction on the ground the evidence had been improperly admitted under section 1107, and the Court of Appeal ruled on that basis, adopting the reasoning of People v. Williams.
In the course of 21 pages of reporter’s transcript, the prosecutor asked only three or four questions concerning recantation and reconciliation.
Family Code section 6211 provides in part:
“ ‘Domestic violence’ is abuse perpetrated against any of the following persons:
“(a) A spouse or former spouse.
“(b) A cohabitant or former cohabitant, as defined in Section 6209.
“(c) A person with whom the respondent is having or has had a dating or engagement relationship.
“(d) A person with whom the respondent has had a child ....
“(e) Child of a party or a child who is the subject of an action under the Uniform Parentage Act....
“(f) Any other person related by consanguinity or affinity within the second degree.”
For purposes of section 6211, “abuse” is defined as “intentionally or recklessly to cause or attempt to cause bodily injury, or sexual assault, or to place a person in reasonable apprehension of imminent serious bodily injury to that person or to another.” (Fam. Code, § 6203.)
“ ‘Cohabitant’ means a person who regularly resides in the household. ‘Former cohabitant’ means a person who formerly regularly resided in the household.” (Fam. Code, § 6209.)
The majority implies the evidence conforms to Darr’s domestic violence paradigm because “Pipes told Deputy Wheeler that defendant had complained about the cleanliness of the apartment on the evening of the assault.” (Maj. opn., ante, at p. 907.) Even assuming this scant evidence, in the abstract, would suffice as tension-building, the majority’s reliance is completely out of context. Pipes testified—and was not contradicted on this record—that she and defendant argued principally over her refusal to pay full rent due to the “unlivable” condition of the house and defendant’s cousin’s failure to make repairs. When, earlier on the night of the incident, the cousin had come to demand payment, Pipes became angry when defendant did not take her side in the dispute. The reference to cleanliness arose in the context of this disagreement. Apparently, defendant had agreed to make some repairs, but complained that he was unable to because the house was too much of a mess. On cross-examination, Pipes agreed with defense counsel that the situation “wasn’t much of an argument, a couple of minutes of getting it off his chest, you getting your complaints off your chest.” Nothing in Pipes’s initial statements to the deputy substantiates a different characterization or reasonably foreshadows an impending cycle of violence.