Kepner’s Rule: Generalized Eyewitness-Identification Expertise Is Ordinarily Admissible Under Iowa Rule 5.702, and Corroboration Is Irrelevant to Admissibility

Introduction

In State of Iowa v. Pat Grant Kepner (Iowa Sup. Ct. Nov. 14, 2025), the Iowa Supreme Court addressed whether a criminal defendant may present expert testimony on the science of eyewitness identification—specifically, generalized testimony explaining how law enforcement identification procedures and human memory processes can affect the reliability and perceived confidence of identifications. The district court excluded the defense expert’s testimony; a jury convicted Kepner of two counts of indecent exposure based largely on the identifications of two eyewitnesses. The court of appeals affirmed. On further review, the Iowa Supreme Court vacated the court of appeals decision, reversed the conviction, and remanded for a new trial.

The opinion clarifies Iowa law in two decisive respects:

Generalized expert testimony on the psychology of eyewitness identification ordinarily satisfies Iowa Rule of Evidence 5.702’s “helpfulness” requirement in cases that rely in whole or part on an eyewitness identification.
Whether eyewitness testimony is corroborated by other evidence is not a gatekeeping factor for admissibility under Rule 5.702.

The decision aligns Iowa with the substantial national consensus favoring admission of properly limited eyewitness-identification expertise, while delineating a clear boundary against “vouching” or case-specific credibility opinions. Justice May authored the unanimous majority (with Justice Mansfield dissenting only on prejudice).

Summary of the Opinion

The Court held that the trial court abused its discretion by categorically excluding the defense expert, psychologist Dr. Kim MacLin, whose testimony was limited to general principles concerning memory, lineup/array administration, and the confidence-accuracy relationship. The Court emphasized:

Such generalized testimony is “sufficiently beyond common experience” to assist jurors under Rule 5.702.
It does not constitute impermissible vouching when it neither evaluates the specific witnesses nor applies the science to the case facts.
Admissibility does not turn on whether the State has corroborating evidence in addition to eyewitness identifications.

On prejudice, the Court applied Iowa’s presumption of prejudice for nonconstitutional evidentiary errors and found the State had not affirmatively established harmlessness. Identity was the only contested issue; the State repeatedly told the jury the case “boils down” to the eyewitnesses; and the standard jury instruction on identification did not substitute for the expert’s specialized education on memory and procedure (e.g., the effect of pre-array instructions, double-blind administration, and post-identification feedback).

Disposition: Decision of Court of Appeals vacated; district court judgment reversed; case remanded for a new trial.

Analysis

Precedents Cited and Their Influence

State v. Galloway (Iowa 1979): A special concurrence representing the majority view effectively adopted a per se exclusion of expert testimony on eyewitness reliability, reasoning that such topics were within jurors’ common knowledge. Kepner recounts Galloway as the former Iowa position but squarely places it in the past.
State v. Schutz (Iowa 1998): Abrogated Galloway’s per se rule and adopted a discretionary approach under Rule 5.702, recognizing a growing national consensus that such expertise can aid jurors. Kepner builds on Schutz, clarifying the “helpfulness” analysis and giving operational content to when and how such testimony should be admitted.
National authorities endorsing admissibility: The Court cites the broad acceptance across 44 states and the federal circuits (see Commonwealth v. Walker, Pa. 2014; State v. Doolin, Iowa 2020 (noting national trend)). Kepner situates Iowa within this modern consensus.
Child sexual abuse expert testimony line: State v. Dudley (Iowa 2014), State v. Jaquez (Iowa 2014), State v. Leedom (Iowa 2020), State v. Payton (Iowa 1992), State v. Montgomery (Iowa 2021), and State v. Allen (Iowa 1997). These cases draw a principled line: experts may educate jurors on general phenomena outside common experience, but may not vouch for truthfulness or indirectly match a complainant’s symptoms to an abuse profile. Kepner imports this template to eyewitness science—general education is permitted; case-specific vouching is not.
Admissibility independent of corroboration: The Court distinguishes Georgia and earlier New York cases that hinged admissibility on lack of corroboration (Johnson v. State (Ga. 2000); People v. Santiago (N.Y. 2011)), noting New York has since moved away (People v. Vaughn (N.Y. 2024)). Kepner adopts the view articulated in State v. Guilbert (Conn. 2012) and In re L.C. (D.C. 2014): corroboration is generally irrelevant to whether such testimony will “help” the jury under Rule 5.702.
Rules and standards: Iowa R. Evid. 5.702 (expert helpfulness), 5.403 (balancing/probative vs. prejudice), 5.103(a) (effect on substantial rights), and harmless-error authorities (State v. Sullivan (Iowa 2004); McGrew v. Otoadese (Iowa 2022); Eisenhauer (Iowa 2019); State v. Buelow (Iowa 2020)) guide the remedy analysis. Kepner reinforces the presumption of prejudice in nonconstitutional evidentiary error when the issue is central and emphasized (Tarbox; Skahill).
Jury instructions: State v. Booth-Harris (Iowa 2020) upheld a version of the ISBA eyewitness instruction. Kepner acknowledges the instruction’s value but explains why it cannot substitute for live expert education on specific scientific phenomena like “chooser” bias and post-identification feedback effects.

Legal Reasoning

The Court’s legal analysis proceeds in four steps:

Helpfulness under Rule 5.702: The Court credits modern social science on memory and identification procedures as “sufficiently beyond common experience” to assist jurors. It highlights topics such as lineup filler selection, simultaneous vs. sequential presentation, double-blind administration, standardized pre-array instructions to avoid “chooser” pressure, post-identification feedback, memory malleability, and the weak correlation between confidence and accuracy. These are not intuitive to lay jurors and thus meet 5.702’s standard.
No vouching or case-specific application: The district court erred by equating general testimony about confidence vs. accuracy or about best practices with impermissible commentary on a specific witness’s credibility. The majority draws directly from the child-abuse line: generalized education is admissible so long as it does not opine on the truthfulness of the particular witness or tailor the science to endorse or undermine a specific identification.
Corroboration is not a gatekeeping factor: The admissibility question turns on whether the testimony will help the jury evaluate the evidence before it, not whether the State has offered additional inculpatory proof. As the Court puts it, corroboration is generally “irrelevant” to the 5.702 analysis. This aligns Iowa with jurisdictions like Connecticut and D.C. and with New York’s recent shift away from a corroboration requirement.
Harmless error and prejudice: Because this is nonconstitutional error, Iowa presumes prejudice unless the record affirmatively shows no effect on substantial rights. Here, identity was the only issue; the State hammered the eyewitnesses as the core of the case in opening and closing; and the generalized jury instruction did not address the specific scientific concepts the expert would have covered. The Court thus cannot say the exclusion was harmless.

Impact

Kepner meaningfully recalibrates Iowa criminal practice regarding eyewitness identifications:

Presumptive admissibility of generalized eyewitness-ID science: In cases where identity is contested and the State relies in any measure on eyewitness identifications, defense (and, potentially, prosecution) experts who provide generalized education without tying opinions to specific witnesses should ordinarily be admitted under Rule 5.702.
Narrower role for “vouching” objections: Trial courts should distinguish impermissible credibility assessment from permissible generalized education. A categorical exclusion on “credibility” grounds is now reversible error if the proffer is properly limited to general principles.
Corroboration no longer blocks admissibility: Prosecutors cannot defeat admissibility by showing other evidence supports guilt. Helpfulness turns on aiding the jury’s evaluation of the eyewitness evidence that is, in fact, before it.
Law-enforcement practices under scrutiny: Expect more litigation and evidence about lineup and show-up protocols. Double-blind administration, proper pre-array instructions, sequential presentation, and filler selection will likely become routine trial topics. Agencies may update policies to match best practices, anticipating expert testimony.
Jury instruction remains helpful but not sufficient: The standard ISBA instruction is not a substitute for scientific education on memory and procedure. Counsel should consider requesting both the instruction and expert testimony.
Appellate posture: Where identity is central and eyewitnesses are highlighted to the jury, erroneous exclusion of such expertise will be difficult to deem harmless.

Complex Concepts Simplified

Generalized vs. case-specific expert testimony: Generalized testimony teaches jurors about scientific principles (e.g., how feedback can inflate confidence). Case-specific testimony applies those principles to say a particular witness is right or wrong. The first is generally admissible; the second often crosses into impermissible vouching.
Double-blind administration: The person showing the array does not know who the suspect is, reducing unconscious cues (“tells”) that can influence the witness.
Sequential vs. simultaneous arrays: Sequential shows photos one at a time, discouraging comparative “which looks closest?” choices. Simultaneous shows all at once. The choice affects error rates.
Filler or lineup bias: If the non-suspect “fillers” don’t match the witness’s description as well as the suspect does, the suspect stands out, increasing the risk of a false identification.
Pre-array “chooser” instruction: Telling a witness that the investigation continues regardless of their selection and that excluding innocents matters can reduce pressure to pick “someone,” which otherwise elevates false IDs.
Post-identification feedback effect: Praise or confirming comments after a selection can inflate a witness’s confidence without improving accuracy.
Confidence vs. accuracy: High confidence at trial may reflect procedural reinforcement rather than genuine memory strength; the two are not reliably correlated absent careful controls.
Vouching: An expert’s direct or indirect assertion that a particular witness is truthful or that their behavior is “consistent with” having witnessed the event; generally barred in Iowa.
Abuse of discretion: A trial court error occurring when legal principles are misapplied or when a decision is based on untenable grounds. Kepner finds misapplication in the categorical exclusion.
Harmless error (nonconstitutional): Iowa presumes prejudice from evidentiary error; reversal is avoided only if the record affirmatively shows no effect on substantial rights. Emphasis on the contested issue and closing arguments inform this analysis.

Practical Guidance for Iowa Practitioners and Courts

For Defense Counsel

Retain a qualified expert in eyewitness-memory science early; prepare an offer of proof that is scrupulously generalized and avoids any case-specific opinions or implicit vouching.
Frame admissibility under Rule 5.702’s helpfulness; cite Schutz and Kepner; emphasize identity centrality and jury education needs.
Anticipate Rule 5.403 challenges by proposing limiting instructions and highlighting how the testimony complements (rather than duplicates) the standard identification instruction.
Use discovery and cross-examination to develop the record on lineup administration: blind vs. non-blind, instructions, sequential vs. simultaneous, filler selection, and any post-ID feedback.

For Prosecutors

Assume generalized identification expertise will be admitted when identity is in dispute; prepare to address it rather than reflexively oppose it.
Bolster lineup practices: adopt double-blind administration, standardized pre-array scripts, careful filler selection, and thorough documentation/video recording of procedures.
Consider your own expert to contextualize best practices and to explain any deviations.

For Trial Judges

Distinguish generalized science (admissible) from case-specific credibility opinions (inadmissible). Avoid categorical exclusions on “credibility” grounds when the proffer is generalized.
Do not condition admissibility on corroboration. Focus on whether the testimony will help the jury evaluate the eyewitness evidence actually presented.
Use Rule 5.403 carefully and case-specifically; when identity is central, the probative value of generalized eyewitness science will ordinarily be substantial.
Consider giving a limiting instruction clarifying that the expert offers general education and that jurors remain the ultimate judges of witness credibility and weight.

The Dissent (Mansfield, J.): A Different View on Harmless Error

Justice Mansfield concurred on the law—agreeing the expert should have been allowed—but dissented on prejudice. In his view, the State’s case was strong regardless of the eyewitnesses’ identifications:

Vehicle evidence: surveillance matched distinctive features of Kepner’s car (silver sedan, sunroof, bent plate, bumper discoloration, rear sag).
License-plate confusion still pointed toward Kepner’s “IAP625.”
Bodycam footage and Kepner’s own interview statements were incriminating; Kepner offered “everything but a direct denial.”
The photo array was of high quality, and defense could have introduced videos of the array administration if helpful.

The dissent also cautioned against routine admission of such experts, citing federal cases that hesitate to require them unless the prosecution’s case rests exclusively on uncorroborated eyewitness testimony (e.g., United States v. Kime; United States v. Villanueva).

Key Takeaways and Significance

Kepner cements Iowa’s “liberal view” of expert admissibility under Rule 5.702 for eyewitness-identification science, aligning Iowa with the national mainstream.
Generalized expert testimony that educates jurors about memory and identification procedures is ordinarily admissible in identity-disputed cases; it does not constitute vouching if it avoids case-specific application.
Corroboration is not a gatekeeper for admissibility; helpfulness to the jury is the touchstone.
Erroneous exclusion in cases centered on identity and featuring emphasized eyewitness testimony will often warrant a new trial given Iowa’s presumption of prejudice.
Law enforcement agencies should expect closer scrutiny of lineup/array protocols and consider updating practices to recognized best standards (double-blind, sequential, standardized instructions, robust documentation).

Conclusion

State v. Kepner marks a significant doctrinal consolidation in Iowa’s treatment of eyewitness-identification expertise. Building on Schutz and the child-abuse expert testimony jurisprudence, the Court draws a bright line between permissible generalized education and impermissible credibility vouching. It further clarifies that corroboration is irrelevant to the admissibility analysis under Rule 5.702. By reversing and remanding for a new trial, the Court underscores the practical importance of allowing jurors to hear the science when identity is the central issue and the State’s case heavily features eyewitness identifications. Kepner will reshape pretrial and trial practices in Iowa, promoting more rigorous lineup procedures and more informed jury deliberations on one of the most consequential and error-prone forms of evidence in criminal cases.