The Court of Appeals, in affirming Keith Johnson’s convictions for armed robbery, aggravated battery, and aggravated assault, upheld the trial court’s grant of the State’s motion in limine regarding expert testimony Johnson proffered on the issue of the reliability of eyewitness identification. Johnson v. State, 236 Ga. App. 252 (5) (511 SE2d 603) (1999). The legal authority for this holding came from Norris v. State, 258 Ga. 889, 890 (1) (376 SE2d 653) (1989), wherein this Court stated that

[t]he determination of a witness’ credibility, including the accuracy of eyewitness identification, is within the exclusive province of the jury. [Cit.] The memory of a witness may not be disparaged by another witness in order to impeach that testimony; it must be done by cross-examination. [Cits.]

We granted the writ of certiorari to consider whether an accused’s rights require that trial courts retain the discretion to admit proffered expert testimony regarding the reliability of eyewitness identifications and whether the Court of Appeals erred in affirming the trial court’s decision to exclude Johnson’s proffer of expert testimony regarding the reliability of eyewitness identifications as a matter of law. We hereby reaffirm our recent holding in Johnson v. State, 271 Ga. 375 (12) (519 SE2d 221) (1999) that the admissibility of expert testimony regarding the reliability of eyewitness testimony is left to the sound discretion of the trial court and disapprove Norris to the extent it can be read as requiring the exclusion of such testimony as a matter of law. Under the facts of this case, however, we find no abuse of the trial court’s discretion in excluding the proffered expert testimony and accordingly affirm the judgment of the Court of Appeals.

[*255] 1. In Jones v. State, 232 Ga. 762 (2) (208 SE2d 850) (1974), we addressed for the first time the admissibility of expert testimony regarding the credibility of eyewitness identification. We held that a witness may not give an opinion about the correctness or incorrectness of the eyewitness identification of the accused because that would invade the province of the jury, id. at 764, so that expert testimony was admissible as to the credibility of a witness only if the subject matter involved organic or mental disorders. Id. at 765. Otherwise, expert testimony should be excluded since the subject matter would be within the scope of the ordinary layperson. Id. Finally, Jones held that expert testimony concerning a witness’ credibility should be excluded where the expert’s opinion is based on a “total lack of or insufficient observation of the assailed witness,” finding inadequate the hypothetical questions posed the witness in that case and noting that the expert “would have been testifying with no knowledge or interview of the eyewitnesses whose testimony was sought to be discredited.” Id.

In our subsequent holding in Norris, supra, 258 Ga. at 890 (1), we upheld the exclusion of proffered expert testimony by the same expert employed by Johnson in the instant case. Although it appears that Norris has been interpreted as excluding expert testimony on eyewitness identification as a matter of law, see, e.g., Cox v. State, 197 Ga. App. 240 (4) (398 SE2d 262) (1990), that interpretation is at odds with our most recent pronouncements in this area. This Court has strongly approved the trial court’s exercise of its discretion in regard to determining the admissibility of this type of expert testimony under the particular facts adduced in each case and has explicitly applied an abuse of discretion standard when reviewing the trial court’s ruling. Johnson, supra, 271 Ga. at 382 (12); Gardiner v. State, 264 Ga. 329 (5) (444 SE2d 300) (1994).

Johnson urges this Court to join the “modern trend” in this area of the law, arguing we should adopt the position taken in United States v. Downing, 753 F2d 1224 (3d Cir. 1985) and People v. McDonald, 690 P2d 709 (Cal. 1984), which hold that it is an abuse of discretion to exclude expert testimony on eyewitness identification when the State’s case against an accused depends wholly upon eyewitness identification, i.e., there is no other substantial corroborating evidence, and when the defense has adduced a detailed offer of proof on the record which explains precisely how the expert’s testimony is relevant to the eyewitness identifications under consideration. ^[1] In[*256] essence, these cases eliminate a trial court’s discretion and mandate the admission of expert testimony where the above factors are present. Our review of the foreign jurisdictions which have addressed this issue in the fifteen years since Downing and McDonald were rendered reveals that a trend has emerged in which the courts have recognized that it may be appropriate to admit expert testimony on the subject of human memory in cases turning on an eyewitness identification. See, e.g., McMullen v. State, 714 S2d 368 (Fla. 1998); Ex Parte Williams, 594 S2d 1225 (Ala. 1992). ^[2] Indeed, only a minority of states and Federal circuits still adhere to the position that expert testimony on eyewitness identification should be excluded as a matter of law. See, e.g., United States v. Smith, 122 F3d 1355 (II) (A) (11th Cir. 1997) (adhering to its ruling in United States v. Holloway, 971 F2d 675, 679 (11th Cir. 1992)); State v. Gaines, 926 P2d 641, 645-649 (Kan. 1996) (expert testimony is “ ‘not the answer to the problems surrounding eyewitness identifications’ ”). However, while the “modern trend” allows the admission of expert testimony on this issue, most foreign courts have rejected the limitation placed on trial courts’ discretion in regard to the admission of expert testimony by the opinions in Downing and McDonald. Rather than pre-determining on an appellate level that qualified, pertinent expert evidence must be admitted in every case where key eyewitness identification is unsubstantiated by other evidence, the modern trend is to allow trial courts to retain their discretion to weigh the admissibility of this evidence under a case-by-case analysis. See, e.g., McMullen, supra, 714 S2d at 370-371. See also 35 AmJur3d, Proof of Facts 1, § 8 (1996). Thus, while the presence of certain factors in a case may strongly favor the admissibility of expert evidence on eyewitness identification, trial courts are not automatically required to admit the evidence; rather, the admissibility of the evidence remains within the trial courts’ control subject to appellate review for abuse of discretion.

Although expert testimony on the reliability of eyewitness identification is excluded as a matter of law in every Federal court in Georgia, United States v. Smith, supra, 122 F3d at 1355, we decline to join the minority of states and Federal circuit courts which have adopted a per se bar to expert testimony on this issue and we disap[*257] prove our holding in Norris, supra, 258 Ga. at 890 (1), to the extent it can be read as mandating the exclusion of such testimony as a matter of law. Likewise, we decline to join the minority of states and Federal circuit courts which have concluded that a trial court necessarily abuses its discretion by refusing to admit qualified, pertinent expert testimony in any case where no substantial evidence exists to corroborate eyewitness identification testimony. Rather, consonant with the position adopted by “[a]n overwhelming majority of both federal and state courts that have addressed this issue,” (footnotes omitted), McMullen, supra, 714 S2d at 370, we adhere to the position followed by this Court in Johnson and Gardiner, supra, and hold that the admission of expert testimony regarding eyewitness identification is in the discretion of the trial court. Where eyewitness identification of the defendant is a key element of the State’s case and there is no substantial corroboration of that identification by other evidence, trial courts may not exclude expert testimony without carefully weighing whether the evidence would assist the jury in assessing the reliability of eyewitness testimony ^[3] and whether expert eyewitness testimony is the only effective way to reveal any weakness in an eyewitness identification. Downing, supra, 753 F2d at 1230-1231, in. 6. However, the admission or exclusion of this evidence “lies within the sound discretion of the trial court, whose decision will not be disturbed on appeal absent a clear abuse of discretion.” O’Neal v. State, 254 Ga. 1, 3 (3) (325 SE2d 759) (1985).

2. We turn now to the facts adduced in Johnson’s case to determine whether the trial court abused its discretion by excluding the testimony of Johnson’s expert, Dr. Steven Cole. ^[4] The crime occurred at an ATM of a bank inside a suburban shopping center. The victim was a 62-year-old white woman who was attacked from behind by a[*258] black male who cut her throat, knocked her to the ground, stabbed her repeatedly and then fled. The attack took approximately one minute and was recorded, at two-second intervals, by a bank video camera. Still photos made from the videotape were introduced into evidence. The victim testified that she spoke to her attacker and remembered his responses to her words. The victim was shown several photo line-ups within a month of the crime and declined to identify any of the displayed individuals as her attacker. Johnson’s photo was not among those shown the victim. Five months after the crime, the victim viewed another photo line-up and selected Johnson’s photo. She identified Johnson at trial as the man who attacked her.

The State did not rely exclusively on the victim’s testimony or the ATM photos. The State also adduced testimony by Brenda Gilpatrick, a white businesswoman, who stated that on the night of the crimes in issue, she had driven to the shopping center to have dinner with a friend in a restaurant at the end nearest to the ATM where the victim was attacked. Ms. Gilpatrick, who had received security training and taught security issues during her 22-year employment in the shopping center industry, noticed a young black man standing on the sidewalk in the area where she intended to walk. The man was fidgeting and looking so intently in her direction that she turned around to see what was interesting the man and realized he was watching the bank area behind her. The man’s behavior was so suspicious that Ms. Gilpatrick decided not to exit her car and instead waited, observing him. The witness stated that it was night and raining at the time, but that the man stood directly under a security light and there was also recessed lighting; the witness described the lighting as “very, very good for a shopping center.” When the man realized that Ms. Gilpatrick was watching him, he “kind of jumped” behind some bushes. At that time, others left the restaurant and Ms. Gilpatrick was able to safely enter and join her dining companion. She immediately raised the matter with her companion; they looked outside, but the man had left.

The following day, after realizing that the attack she heard about on the news had occurred at the ATM near the restaurant, Ms. Gilpatrick contacted the police. She viewed the ATM photos of the attacker and told the police she was confident it was the same person she had seen the previous evening. She then provided police with a detailed description of the man she had observed, including details about his eyebrows and the manner in which his right ear stuck out and his left ear looked pinned back in a “curled effect.” That same day Ms. Gilpatrick spent two hours working with a police artist to create a composite picture of the man she had observed, describing the picture as a “nine” in comparison to the man she remembered. Ms. Gilpatrick was also shown several photo line-ups within a month[*259] of the crime and did not find the man she had seen in the line-ups, though she used one photo to describe in detail why that individual was not the correct one by pointing out to the detective the differences in nose thickness, eyebrows, hairline and other features. Five months later Ms. Gilpatrick was shown the photo line-up containing Johnson’s photo. ^[5] The detective who handled the matter testified that he had not finished sitting down after handing Ms. Gilpatrick the line-up before she had selected Johnson’s photo. Ms. Gilpatrick identified Johnson at trial as the man she saw in the shopping center.

In addition to the eyewitness’ identification, the State introduced evidence of a similar transaction, in which a 60-year-old white man was robbed at gunpoint at an ATM located in a shopping center. Although the victim of this crime was not able to identify his attacker, the victim was a weapons hobbyist who recognized the type of weapon carried by the attacker (a “West German import R. G. Ten, twenty-two caliber”) and the victim was able to get the license plate of the car in which the attacker left. Police discovered that the car had been purchased for Johnson by his uncle and located the car, with Johnson behind the wheel and an R. G. .22 caliber handgun on the floorboard, shortly after the crime.

In his statement to police, Johnson stated that he was a crack cocaine addict and “that he had basically no memory or recollection of the months of December and thereafter because of his addiction.” Relatives of the mother of Johnson’s child testified that at least part of the evening in issue he was at a birthday party for the child’s great grandmother, though testimony by the child’s grandfather cast doubt on whether that party occurred in 1993, at the time of the crimes when the child was barely a month old, or the next year, when the child was a one-year-old.

We have set forth the above evidence in considerable detail in order to discuss why the exclusion of the proffered testimony by Johnson’s expert was not an abuse of discretion in this case. The transcript reveals that Johnson’s expert, Dr. Cole, testified generally about what science has revealed regarding the manner in which images are processed and retained by the brain and discussed several particulars affecting an individual’s ability to identify another, including matters well within the knowledge of an average jury such as the nature of lighting and weather conditions existing at the time the person is viewed, the influence other sources such as photos could have on identification, and the effect the passing of time has on a witness’ accuracy. Dr. Cole also discussed factors less likely to be fully[*260] understood by jurors, such as the detrimental effect acute stress can have on identification; difficulties in cross-racial identifications; the likelihood that a victim will focus on an attacker’s weapon rather than the attacker’s face; and studies indicating that a greater level of confidence a witness has in his identification does not always mean a greater accuracy in identification.

When this expertise is applied to the facts adduced by the State, particularly the testimony of Brenda Gilpatrick, it is apparent that no abuse of the trial court’s discretion occurred in the exclusion of Dr. Cole’s testimony. Ms. Gilpatrick was not under acute stress and was not faced with a deadly weapon at the time she viewed Johnson. As to cross-racial misidentification, the transcript is clear that Ms. Gilpatrick did not identify Johnson based on generalized racial characteristics but on individual details such as the thickness of his eyebrows and neck, the location of his hairline, and the shape of his ears. As to the certainty of Ms. Gilpatrick’s identification, the trial court admitted into evidence the composite drawing created less than 24 hours after Ms. Gilpatrick viewed Johnson in the shopping center; the trial court thus knew the jury would be able to compare the composite to Johnson’s appearance at trial as well as to photos of Johnson admitted into evidence which depicted his appearance around the time of the crimes. Furthermore, at the time the trial court excluded the proffered testimony of Dr. Cole, the court had heard the testimony by the victim and Ms. Gilpatrick, the extensive cross-examination of both eyewitnesses by defense counsel, ^[6] and the State’s similar transaction evidence.

We find that while the testimony of Dr. Cole might have been helpful to some degree, under the circumstances in this case there was no clear abuse of discretion in the trial court’s refusal to admit Dr. Cole’s testimony. Accordingly, because the grant of the State’s[*261] motion in limine regarding expert testimony on the reliability of eyewitness identification did not constitute error, we affirm the judgment of the Court of Appeals.

Judgment affirmed.