At issue in this case is the admissibility of expert testimony regarding the battered woman syndrome when offered to assist the jury in understanding the complainant’s testimony and actions. We hold that expert testimony regarding *580the battered woman syndrome is admissible only when it is relevant and helpful to the jury in evaluating a complainant’s credibility and the expert witness is properly qualified.
Generally, battered woman syndrome testimony is relevant and helpful when needed to explain a complainant’s actions, such as prolonged endurance of physical abuse accompanied by attempts at hiding or minimizing the abuse, delays in reporting the abuse, or recanting allegations of abuse. If relevant and helpful, testimony regarding specific behavior is permissible. However, the expert may not opine whether the complainant is a battered woman, may not testify that defendant was a batterer or guilty of the instant charge, and may not comment on the complainant’s truthfulness. Moreover, the trial court, when appropriate, may preclude expert testimony when the probative value of such testimony is substantially outweighed by the danger of unfair prejudice.
In this case, the expert testimony was arguably relevant and helpful in understanding complainant’s actions in tolerating physical abuse over a period of years. Moreover, it may have been relevant in explaining why complainant did not report similar incidents earlier. On the other hand, its relevance did not reach the level found in other battered women cases that have considered this issue.1 Complainant did not remain in the relationship until the date of the assault and try to hide or deny the abuse, did not delay reporting this incident, and did not later retract the claim of abuse. Instead, complainant testified that the relationship ended one month before the assault, explained that she immediately reported the sexual assault, and has consistently maintained that the *581abuse occurred. Although the testimony was arguably relevant and helpful, on these facts, we are persuaded that a more direct connection and factual premise is necessary, and, hence, we deem the trial court’s decision to admit the testimony to be error.
Nonetheless, we deem the error harmless in light of the limited nature of the testimony and the other physical and testimonial evidence of abuse. The expert merely explained the characteristics of a battered woman. He neither testified that complainant’s behavior was consistent with such traits, nor opined about complainant’s truthfulness or whether complainant was a battered woman. Combining the physical evidence of sexual abuse with complainant’s testimony, we are persuaded that the limited nature of the expert testimony could not have affected the jury’s decision to convict. Accordingly, we reverse the decision of the Court of Appeals with respect to admission of this expert testimony, but affirm the result because of the harmless nature of the testimony.
i
Defendant was charged with first-degree criminal sexual conduct,2 breaking and entering an occupied dwelling with intent to commit criminal sexual conduct,3 and breaking and entering an occupied dwelling with intent to commit larceny.4 A jury acquitted defendant of both breaking and entering charges, but convicted him of first-degree criminal sexual conduct. The trial court sentenced defendant to fifteen to twenty-five years in prison.
*582For several years before this incident, defendant and complainant shared an on-again, off-again romantic relationship.5 6 At one point, they lived together with defendant’s mother and later moved into their own apartment. However, this latter arrangement ended after about a week, on December 13, 1989, when heated arguments and physical abuse compelled complainant to ask defendant to move out.6 Complainant testified that this ended their relationship, whereas defendant claimed that the relationship continued nonetheless until the date of this alleged assault. At the time of the assault, complainant was six months pregnant with defendant’s child.
At trial, complainant testified that their relationship began well, but later progressed into both verbal and physical abuse. Defendant apparently became extremely jealous of complainant; he accused her of dating other men and became angered at any intimation that she was looking at other men in person, on television, or in pictures. This jealously often turned to rage and beatings, followed by compelled sexual intercourse. Complainant described this behavior as being a "game” for defendant. Indeed, a neighbor confirmed the rage and persistent behavior, testifying that defendant frequently came to complainant’s apartment demanding to see her, and that she would refuse to let him in and tell him to leave. On many occasions, the neighbor observed defendant lurking in the laundry room waiting for complainant to arrive home.
On the afternoon of January 23, 1990, complain*583ant testified that defendant entered her apartment by breaking , the chain lock on the door.7 After verbal abuse, he proceeded to slap her with an open hand and eventually to rape her. The rape occurred after defendant grabbed complainant by the neck and forced her face into a pillow while stating: "this is what you want, this is what you are going to see your boyfriend for, I can do it better than he can.” In doing so, he scratched and left "hickeys” on the back of her neck. As a result of the intercourse, complainant endured premature labor contractions and accompanying pain. Despite knowing of the contractions, defendant immediately left the apartment after the incident.8 Complainant then drove herself to the hospital.
When complainant arrived at the hospital, she was crying and had bruises and scratches on her body and dried blood on her lip.9 She was treated for premature labor and remained in the hospital overnight. Doctors and nurses performed' a rape test and noticed that her vaginal area was red and swollen. Thereafter, hospital personnel notified the police, resulting in complainant’s consultation with Sergeant Fred Reid regarding her allegation of rape.
When released from the hospital, complainant stopped at her apartment to retrieve some clothing *584and drop off some material regarding the police investigation. She then went to stay with her mother. The following day complainant returned to her apartment and noticed that the damaged chain lock was missing, along with the pillow case used to muffle her screams, two business cards from Sergeant Reid, and an audio tape from the answering machine, which allegedly contained harassing messages from defendant. In an open cookbook, however, she found several written notes left by defendant describing complainant as a liar, a cheat, and other more graphic and profane names. He also made threatening remarks like "make my morning” and questioned how she could accuse him of rape. At trial, the cookbook was admitted into evidence. Defendant admitted writing these notes and taking the business cards,10 but denied taking anything else.
During cross-examinations and defendant’s presentation of his case, complainant was portrayed as a liar, a perjurer, a self-mutilator, and an embezzler. Defendant produced several witnesses who allegedly saw defendant and complainant together on New Year’s Eve, 1989, and several family members who saw them together after December 13, 1989, the date complainant said the relationship had ended. Defense counsel also elicited from complainant instances in which she engaged in self-mutilation, e.g., burning herself with a cigarette, in order to get defendant’s atten*585tion.11 Defendant’s theory was that complainant pursued the rape charge in retaliation for threats that defendant would report complainant’s embezzlement from her employer.
To rebut some of these claims and to help evaluate complainant’s credibility, the prosecution called in its case in chief a clinical psychologist, Dr. Louis Oken, who was trained in the field of domestic violence and the battered woman syndrome. Over defendant’s objection, Dr. Oken described the syndrome and explained that women often remain in a relationship with a boyfriend or husband even though abuse is occurring. The syndrome develops in stages, beginning with minor beatings and progressing to more severe beatings, followed by a period of nonviolence. During this period, the battered woman begins to deny, repress, or minimize the abuse rather than be outraged. He noted, however, that people react differently depending on one’s well-being and living situation, thus making it difficult through initial consultations to determine whether a person is in a battering situation.
Because he had never treated complainant or defendant, Dr. Oken limited his testimony to generalities associated with the battered woman syndrome. Accordingly, he did not indicate whether complainant’s behavior was consistent with a battered woman, did not give an opinion whether this particular complainant suffered from the battered woman syndrome or that defendant was a batterer, did not opine whether he thought complainant was being truthful, and did not comment on defendant’s guilt.
In closing arguments, both sides stressed the credibility of its witnesses. Defense counsel said *586complainant was not credible and that the intercourse was consensual, whereas the prosecution contended that she was credible, that her claims were consistent with physical evidence of sexual assault, and that uninterested medical personnel relayed this information. With respect to Dr. Oken, the prosecutor theorized that from complainant’s testimony and Dr. Oken’s rendition of the syndrome
you have to be convinced that she, in fact, was a battered woman, and falls into the criteria that he gave you. She went back and tried to work out that situation after sustaining numerous beatings, and you can’t understand that, and I can’t understand that, but he tells you that it is not a reasonable conclusion, but it is something that women do all the time.
After conviction, defendant sought review in the Court of Appeals, which affirmed the conviction. Unpublished opinion per curiam of the Court of Appeals, issued December 9, 1993 (Docket No. 140721) (Michael J. Kelly, J., dissenting). The Court of Appeals majority held that the expert testimony was properly admitted in order to assist the trier of fact in understanding the evidence.12 "While the average person without special knowledge might reject the complainant’s testimony that she remained in an abusive relationship off and on for over a year as incredible, the expert testified that it is actually quite common for people who *587suffer physical abuse at the hands of someone they love to continue in relationships with their abusers.” Slip op at 1.
On December 29, 1994, this Court granted leave to appeal limited to the issue whether evidence of the battered woman syndrome was admissible.13
ii
Before permitting expert testimony, the trial court must find that the evidence is from a recognized discipline, as well as relevant and helpful to the trier of fact, and presented by a witness qualified by "knowledge, skill, experience, training, or education . . . .” MRE 702; People v Beckley, 434 Mich 691; 456 NW2d 391 (1990). On request, the trial judge may deem a limiting instruction appropriate and, in certain cases, may exclude the testimony because the probative value of the syndrome evidence is substantially outweighed by the danger of unfair prejudice. MRE 403; Beckley, supra at 725 (opinion of Brickley, J.); id. at 741 (opinion of Boyle, J.). On appeal, our duty is to review the decision to admit such testimony for an abuse of discretion.14
A
The expert testimony in this case involved what has become known as the battered woman syndrome. Dr. Lenore Walker in her book, The Battered Woman (New York: Harper & Row, 1979), brought to the forefront this concept and explanation of abuse between husband and wife and boy*588friend and girlfriend. Dr. Walker defined a battered woman as
a woman who is repeatedly subjected to any forceful physical or psychological behavior by a man in order to coerce her to do something he wants her to do without any concern for her rights. Battered women include wives or women in any form of intimate relationships with men. Furthermore, in order to be classified as a battered woman, the couple must go through the battering cycle at least twice. Any woman may find herself in an abusive relationship with a man once. If it occurs a second time, and she remains in the situation, she is defined as a battered woman. [Id. at XV.]
Dr. Walker notes that a battering or abusive relationship often results in criminal litigation either against the abuser or the abused who retaliates. When the precipitating facts of the syndrome are offered into evidence in either case, the syndrome is not easily conceptualized by lay persons.15 Accordingly,
[e]xpert testimony on the battered woman syndrome would help dispel the ordinary lay person’s perception that a woman in a battering relationship is free to leave at any time. The expert evidence would counter any "common sense” conclusions by the jury that if the beatings were really that bad the woman would have left her husband much earlier. Popular misconceptions about battered women would be put to rest, including the beliefs that the women are masochistic and enjoy the beatings and that they intentionally provoke their husbands into fits of rage. [State v Hodges, 239 Kan 63, 68-69; 716 P2d 563 (1986), summarizing The Battered Woman, supra at 19-31.]
*589In most cases, the battered woman syndrome is offered by the defendant in a case of homicide in which the defendant is claiming self-defense. As one court has explained:
[E]xpert scientific evidence concerning "battered-woman’s syndrome” does not aid a jury in determining whether a defendant had or had not behaved in a given manner on a particular occasion; rather, the evidence enables the jury to overcome common myths or misconceptions that a woman who had been the victim of battering would have surely left the batterer. Thus, the evidence helps the jury to understand the battered woman’s state of mind. [State v J Q, 130 NJ 554, 574; 617 A2d 1196 (1993).]
Although we do not express approval or disapproval of this use, we note that our Court of Appeals recently recognized that a majority of jurisdictions favor the admissibility of expert testimony on the issue of the battered woman syndrome when offered as a means of self-defense. See People v Wilson, 194 Mich App 599, 603; 487 NW2d 822 (1992), citing anno: Admissibility of expert or opinion testimony on battered wife or battered woman syndrome, 18 ALR4th 1153.
The instant case, however, falls in the minority of situations in which the evidence is offered to help evaluate the credibility of the complainant instead of exculpating the accused. Use of expert testimony in this situation is an issue of first impression for this Court. Nonetheless, our recent decision in Beckley, supra, and decisions from our sister jurisdictions guide our analysis.
B
In Beckley, this Court addressed whether expert testimony regarding the rape trauma syndrome is *590admissible in child sexual abuse cases in order to rebut the inference that the victim’s behavior was inconsistent with that of an actual sexual abuse victim. In a plurality opinion, we held that this expert testimony is generally admissible when the scientific or technical evidence is from a recognized discipline, the testimony is helpful to the trier of fact in understanding relevant evidence, and the expert is qualified. Id. at 711 (opinion of Brickley, J.); id. at 736-737 (opinion of Boyle, J.). Assuming these tests are satisfied, the expert may testify regarding the characteristics of the syndrome and whether the complainant’s behavior is consistent with those traits.16
However, seven justices agreed that syndrome evidence is not admissible to demonstrate that abuse occurred. Id. at 724 (opinion of Brickley, J.); id. at 734 (opinion of Boyle, J.); id. at 744 (opinion of Archer, J.). The Court also agreed that the expert may not give an opinion about whether the complainant is being truthful or the defendant is guilty. Moreover, five justices agreed that where syndrome evidence is merely offered to explain certain behavior, the Davis/Frye17 test for recognizing an admissible science is inapplicable.18 Beckley, supra at 721, 734.
The basis for the three separate opinions in Beckley stemmed from disagreement regarding the necessary foundation for and the parameters of this expert testimony. Justice Brickley19 would limit its admission " 'for the narrow purpose of rebutting an inference that a complainant’s post-incident behavior was inconsistent with that of an *591actual victim of sexual abuse, incest or rape.’ ”20 Id. at 710 (citation omitted). Justice Archer21 concurred in part, but would hold that an expert only can testify in generalities and cannot discuss whether the victim’s behavior is consistent with that of other abuse victims. Id. at 744. On the other hand, Justice Boyle22 would allow the expert to testify about these similarities to assist the jury in deciding a fact at issue. Id. at 736.
c
Today, we extend the general holding in Beckley to expert testimony of the battered woman syndrome so that the expert may, when appropriate, explain the generalities or characteristics of the syndrome.23 We emphasize, however, "that the admissibility of syndrome evidence is limited to a description of the uniqueness of a specific behavior brought out at trial.” Beckley, supra at 725 (opinion of Brickley, J.). In other words, we do not adopt the battered spouse syndrome, but will permit testimony regarding specific behavior where relevant and helpful to the factfinder. Furthermore, we extend the prohibitions agreed on by seven justices in Beckley — the expert cannot opine that complainant was a battered woman, may not testify that defendant was a batterer or that he is guilty of the crime, and cannot comment on whether complainant was being truthful.24
*592However, finding that expert testimony is generally admissible on the issue of the battered woman syndrome does not obviate the threshold determinations for every trial court — helpfulness and relevancy. Thus, we turn to the confines of its admissibility and whether the syndrome evidence in this case was relevant and helpful to the jury. We note that defendant does not seriously contest that Dr. Oken was a qualified expert,25 and he concedes, and we agree, that battered woman syndrome evidence is from a recognized discipline.26
D
Generally, expert testimony is needed when a witness’ actions or responses are incomprehensible to average people. This may include, for example, when a complainant endures prolonged toleration of physical abuse and then attempts to hide or minimize the effect of the abuse, delays reporting the abuse to authorities or friends, or denies or recants the claim of abuse.27 Schroeder, Using *593battered woman syndrome evidence in the prosecution of a batterer, 76 Iowa LR 553 (1991). Only when those or similar facts are at issue and expert testimony would be helpful in evaluating a witness’ testimony is it permissible to admit battered woman syndrome evidence in the prosecution’s case in chief.28
As one of the first jurisdictions to consider the use of the battered woman testimony against a defendant, the New Hampshire Supreme Court in State v Baker, 120 NH 773; 424 A2d 171 (1980), found it relevant and admissible in order to rebut the defendant’s evidence that he was insane at the time he attempted to murder his wife. The court found the evidence probative because it offered "an alternative explanation for the defendant’s assault on his wife.” Because the testimony directly refuted defendant’s insanity, theory, it was relevant and could not have confused or misled the jury. Id. at 775.
In the decade since Baker, courts have limited expert testimony to situations in which the complainant endured prolonged toleration of abuse, but hid or denied that it occurred, delayed reporting the underlying assault, or recanted after alleg*594ing the abuse to have occurred.29 For example, in State v Ciskie, 110 Wash 2d 263; 751 P2d 1165 (1988),30 the defendant was charged with four counts of raping his girlfriend. At trial, the defendant claimed that the complainant received bruises and lacerations from minor accidents. The defendant also asserted that if he was really hürting her she could have immediately sought medical attention, called the police, or ended the relationship at any time instead of months later. The explicit premise of the defense was that shé was a liar and that her behavior was inconsistent with that of an actual abuse victim.
To explain the victim’s failure to leave or immediately report these abuses, the trial judge permitted, and the Washington Supreme Court affirmed, the admission of expert testimony regarding the battered woman syndrome. "Neither logic nor law requires us to deny victims an opportunity to explain to a jury, through a qualified expert, the reasons for conduct which would otherwise be beyond the average juror’s understanding.”31 Id. at 265.
*595A similar need for expert testimony arose in Arcoren v United States, 929 F2d 1235 (CA 8, 1991), in which the complainant recanted her story four months later at trial. Immediately after the rape, the complainant flagged down a police car and reported the rape, recounted the same story to nurses, doctors, and criminal investigators, and, three days after the assault, gave a sworn statement before a grand jury. In what the United States Court of Appeals for the Eighth Circuit described as a "bizarre situation,” it held32 that
[a] jury naturally would be puzzled at the complete about-face she made [at trial], and would have great difficulty in determining which version of [the complainant’s] testimony it should believe. If there were some explanation for [the complainant’s] changed statements, such explanation would aid the jury in deciding which statements were credible. . . . [E]xpert testimony regarding the battered woman syndrome provided that explanation to the jury. [Id. at 1240.]
Likewise, in State v Frost, 242 NJ Super 601; 577 A2d 1282 (1990),33 the defendant had a three and one-half year relationship with the complainant, interrupted by the defendant’s imprisonment on theft charges. When the defendant was re*596leased, he returned to the complainant’s residence, believing that she was responsible for his conviction. He beat her, cut her arm, and raped her. She initially denied the abuse and told her mother not to call the police. After the victim was treated for a laceration on her arm, she went to spend the day with the defendant and their baby. She also visited the defendant while in jail pending trial.
At trial, the defendant testified that the intercourse was consensual and that the complainant was lying. The court found the expert testimony admissible: "Having heard that she spent the entire day with defendant, the jury quite properly could have concluded — based on the misconception that battered women are free to leave — that, if [the complainant] had indeed been assaulted that day, she would have immediately called for help.” Id. at 614.
Finally, in State v Borrelli, 227 Conn 153; 629 A2d 1105 (1993), the court held expert testimony admissible to explain why the complainant might have immediately reported the crime and then recanted the allegation at the hearing on the motion to dismiss and at trial. She testified at trial that she tied the defendant’s hands and feet rather than he tying hers, and that her initial allegations were offered in order to get the defendant arrested so he would be forced to undergo drug treatment. The court noted that the expert testimony "offered a different explanation, one beyond the knowledge and understanding of the average juror — that the statement was true, and the victim’s recantation was a pattern of typical behavior consistent with battered woman’s syndrome.”34 Id. at 171._
*597E
In the instant case, while there may have been some basis for the testimony, we find the necessary factual underpinnings for admission of expert testimony lacking. Certainly there may have been a question why complainant tolerated prolonged abuse without reporting it to authorities or friends. However, defendant never denied that some abuse occurred. Furthermore, complainant testified that the relationship ended one month before the assault and did not attempt to hide or deny the instant sexual assault. Moreover, complainant did not delay reporting this incident, but, instead, immediately sought medical attention with accompanying discussions with police. Two days later, a formal complaint was issued. Complainant also never recanted that the assault occurred.
The prosecution’s contention that she remained in the relationship in spite of the abuse does not by itself make it relevant and helpful to a material issue. Indeed, this contention is belied by complainant’s own testimony that the relationship ended one month before the incident. Expert testimony usually is not needed to explain alternative prosecution theories, but to explain things not readily comprehensible to an average juror. Because complainant has consistently maintained that the relationship ended in December and there is no evidence that complainant hid or minimized, delayed reporting, or recanted the abuse, we reject the prosecution’s contention that the battered woman syndrome was relevant in this case.
On this record, it was questionable whether complainant’s testimony and actions would be incomprehensible to average people so that admission of expert testimony was warranted. Our re*598view of the facts leads us to conclude that it was error to admit this testimony. We sympathize with trial judges who face the difficult task whether to admit such testimony in cases where the factual underpinnings are not always clear and identifiable.35 However, while a bright-line test may not be possible, we believe the foundational confines provided by our sister jurisdictions accurately limit the use of this testimony. Even if there are other permissible uses for the testimony, we are persuaded that the instant case is not one of them. Accordingly, we find that the trial judge abused his discretion in admitting the testimony.
hi
Although it was error to admit Dr. Oken’s testimony, we find the error harmless in light of other physical evidence of abuse, complainant’s testimony, and the limited nature of Dr. Oken’s offering. Dr. Oken merely described the basic characteristics of the syndrome and the different stages of development. He explained various responses, reactions, and methods of coping. He also indicated that there is no one set of criteria to determine whether a woman is being battered. Indeed, in his practice of counseling, he finds it difficult to determine whether a woman is being battered or the defendant is a batterer even after a number of visits, and that it is impossible to determine after just one visit. In making this observation, he never opined whether complainant’s behavior was consistent with this general criteria and, in fact, indi*599cated that he never met or treated complainant or defendant. Moreover, he explicitly disclaimed any intimation that complainant was being truthful or that she was a battered woman. While the testimony was lengthy, the substance and application were minimal.
The other testimony, however, revealed more than just a credibility contest. Physical evidence and testimony by medical personnel confirmed that some abusive activity occurred and that defendant was enraged with complainant. Doctors and nurses testified that complainant was crying when she entered the hospital, that she had bruises on her face and body, scratches and hickeys on her neck, and a red and swollen vaginal area. The theory that the injuries were self-inflicted was dismissed in part by the medical personnel who opined that the angle of the scratches could not have been self-inflicted.
Additionally, the cookbook containing notes by the defendant and acknowledged by him at trial corroborated complainant’s testimony of verbal abuse and rage toward complainant. Moreover, the testimony from complainant’s neighbor regarding defendant’s persistence and accompanying rejection by complainant undercut the possibility of consensual intercourse on January 23, 1990. Coupling these facts with complainant’s testimony, Dr. Oken’s testimony could not have affected the jury’s decision to convict. In other words, the limited nature of the testimony could not have given the jury the "much sought-after hook on which to hang its hat.” Beckley, supra at 722 (opinion of Brickley, J.).
Moreover, the prosecutor’s closing statement, which invited the jury to believe that complainant was a battered woman on the basis of Dr. Oken’s testimony, was not error in itself. Prosecutors are *600permitted to argue the evidence and make reasonable inferences in order to support their case theory. People v Bahoda, 448 Mich 261, 282; 531 NW2d 659 (1995). However, given our holding with respect to the testimony itself, we review its effect on the jury’s verdict. In this case, we are persuaded that it could not have affected the jury’s decision. Dr. Oken testified that he could not even determine whether a person was a battered woman after one visit, let alone without ever treating complainant or defendant. Given that the prosecutor’s contention was not evidence36 and was based on Dr. Oken’s general explanation of the syndrome, it could hardly have been enough to tip the scale in favor of conviction.
iv
At issue in this case is the admissibility of expert testimony regarding the battered woman syndrome when offered to explain a complainant’s testimony and actions. We hold it admissible only when the witness is properly qualified and it is relevant and helpful in understanding an issue in the case. Typically, this includes an explanation of the complainant denying or minimizing the abuse, delays in reporting, or subsequently recanting the abuse.37
In this case, we deem the expert testimony irrelevant and not helpful in explaining any fact in issue. Hence, the trial judge abused his discretion in admitting the testimony. Nonetheless, the limited nature of the testimony, coupled with *601physical and corroborating evidence of abuse, makes the error harmless. Hence, we reverse the Court of Appeals decision with respect to the expert testimony, but affirm the result because of the harmless nature of the testimony.
Brickley, C.J., and Boyle, Mallett, and Weaver, JJ., concurred with Riley, J.