by Douglas Ankney

The Supreme Court of New Jersey unanimously held that a trial court confronted with a substantive challenge to expert fingerprint testimony must assess the evidence’s reliability before admitting it and may not discharge its gatekeeping obligation by relying on a century of judicial acceptance. Because the defendant supported his challenge with two national reports raising questions about latent print analysis, the Court agreed with the Appellate Division that the trial court should have conducted a pretrial hearing. The Court declined to reverse French G. Lee’s conviction, appointed a Special Adjudicator to develop a full expert record, expressed no view on admissibility, and retained jurisdiction.

Background

A Moorestown restaurant was burglarized twice in September 2018. After the first burglary, police lifted one latent print from a pizza oven. After the second, they lifted four from the cash register. The prints were submitted to the Automated Fingerprint Identification System (“AFIS”), which identified Lee as the suspected source. Lieutenant Michael Wiltsey of the Burlington County Prosecutor’s Office then applied the four-step ACE-V method (Analysis, Comparison, Evaluation, and Verification) and concluded that all five impressions originated from the same source as Lee’s prints. Nothing else linked Lee to the burglaries.

Before trial, Lee moved to exclude the fingerprint evidence, relying principally on a 2009 National Academy of Sciences report (“NAS Report”) and a 2016 report of the President’s Council of Advisors on Science and Technology (“PCAST Report”). He argued that the reports raised concerns about examiner consistency, the discipline’s subjectivity, the absence of objective or uniform identification criteria, black-box error rates of one in 306 and one in 18, assumptions that fingerprints are unique and unchanging, confirmation bias, and limited empirical testing. The PCAST Report nevertheless characterized latent print analysis as “a foundationally valid subjective methodology,” while warning that its false-positive rate was substantial and likely higher than many jurors would expect.

The State responded that the reports did not justify departing from nearly a century of New Jersey precedent admitting fingerprint evidence, beginning with State v. Cerciello, 86 N.J.L. 309 (E. & A. 1914). It argued that Frye v. United States, 293 F. 1013 (D.C. Cir. 1923), continued to govern because Lee had not shown a change in the underlying science and cited federal decisions finding ACE-V sufficiently reliable. The trial court denied the motion without an evidentiary hearing, relying on historical acceptance, those federal rulings, and the absence of a decision declaring ACE-V unreliable. The parties had agreed that no hearing under N.J.R.E. 104 was required.

A jury convicted Lee of two counts of third-degree burglary, and the court imposed an aggregate six-year extended term with two years of parole ineligibility. The Appellate Division reversed. It held that the failure to conduct a pretrial reliability hearing was independently reversible error, that the court abused its discretion by declining to question prospective jurors about fingerprint evidence, and that testimony interpreting the surveillance videos violated State v. Watson, 298 A.3d 1049 (N.J. 2023). It concluded that the cumulative effect of the errors denied Lee a fair trial. The Supreme Court granted the State’s petition for certification.

Analysis

The Court began its analysis by observing that N.J.R.E. 702 permits qualified expert testimony when scientific, technical, or other specialized knowledge will assist the trier of fact. The proponent must establish that the subject is beyond the average juror’s ken, that the science is sufficiently reliable, and that the witness has adequate expertise. State v. Olenowski (“Olenowski I”), 289 A.3d 456 (N.J. 2023). Only the second requirement was disputed, and the State bore the burden of proving reliability. State v. Nieves, 345 A.3d 1127 (N.J. 2025); State v. Cassidy, 197 A.3d 86 (N.J. 2018).

Until 2023, New Jersey criminal courts assessed reliability under Frye’s general-acceptance standard, the Court stated. In In re Accutane Litig., 191 A.3d 560 (N.J. 2018), the Supreme Court clarified for civil cases that reliability turns on the methodology and reasoning underlying the proposed testimony, not general acceptance alone. Olenowski I adopted a similar approach for criminal cases. The Court explained that courts now examine “the soundness of the methodology used to validate a scientific theory or technique, the strength of the reasoning underlying it, and the accuracy of the theory or technique in practice,” guided by the nonexclusive Daubert factors, i.e., testing, peer review and publication, error rate, and general acceptance. Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 579 (1993). However, New Jersey has not adopted the full body of Daubert case law, according to the Court.

Olenowski I left prior Frye-based rulings undisturbed but instructed that evidence previously approved should be reassessed under the current standard when its underlying scientific reliability has changed. The Court stated that State v. Olenowski (“Olenowski II”), 304 A.3d 598 (N.J. 2023), reinforced that expert methodology should not be “frozen in time” when new research calls earlier precedent into question.

The Gatekeeping Obligation

The Court noted that trial courts serve as gatekeepers and must determine that expert testimony is sufficiently reliable before it can be presented to a jury. State v. J.L.G., 190 A.3d 442 (N.J. 2018). That role must be rigorous because it protects jurors from “unsound science through the compelling voice of an expert.” Accutane. When a party challenges expert testimony under N.J.R.E. 702, the court should conduct an admissibility hearing under N.J.R.E. 104, although the decision whether to hold one ordinarily rests within the court’s discretion. State v. J.R., 152 A.3d 180 (N.J. 2017); State v. Cain, 133 A.3d 619 (N.J. 2016).

Lee directly challenged the proposed testimony with specific concerns drawn from reports by two reputable national bodies. Without resolving those concerns, the Court held that they raised “legitimate issues that warrant further evaluation.” The trial court therefore had to evaluate reliability rather than rely on historical practice and earlier federal rulings, the Court determined.

All four federal appellate decisions cited by the State predated the PCAST Report, and two predated the NAS Report, so the Court reasoned they could not account for later developments, including black-box error rates. It also observed that no published New Jersey decision had examined fingerprint reliability in depth or reviewed an evidentiary hearing on the subject and that the extent to which courts elsewhere had examined the newer concerns remained unclear.

The parties’ agreement not to request a hearing neither resolved the contested reliability question nor constituted invited error under State v. A.R., 65 A.3d 818 (N.J. 2013). Both sides had presented opposing positions and sought a ruling on the merits. As gatekeeper, the trial court remained obligated to assess reliability. The Court nevertheless recognized the difficult position in which the parties placed the trial court and advised defense counsel to coordinate with the Public Defender’s forensic science unit and county prosecutors to consult the Attorney General when similar challenges arise.

The Special-Adjudicator Hearing

Because the NAS and PCAST reports raised questions requiring evaluation on a developed record, the Court ordered a plenary hearing before a Special Adjudicator, following approaches used in State v. Henderson, 27 A.3d 872 (N.J. 2011), and State v. Moore, 852 A.2d 1073 (N.J. 2004). The Special Adjudicator must determine whether the fingerprint evidence presented at Lee’s trial satisfies Olenowski I and, if it does, whether limitations or guardrails should govern its admission and what they should be. The Court also instructed the Special Adjudicator to consider whether revised model jury charges are needed and to submit written findings and conclusions. The parties and existing amici may present testimony and other evidence, including expert testimony. The Court stressed that it expressed no view on the hearing’s outcome.

The Court deferred the voir dire and video-narration issues pending the fingerprint-reliability hearing because admissibility of the fingerprint evidence was the threshold question.

Conclusion

The Court declined to reverse Lee’s conviction at this stage. Because the fingerprints were the only evidence linking him to the burglaries, the conviction cannot stand if the Court ultimately finds the evidence unreliable. If the evidence is admissible, the Court will address any remaining relevant issues.

Accordingly, the matter was summarily remanded for the Special Adjudicator’s hearing, and the Court retained jurisdiction. See State v. Lee, 2026 N.J. LEXIS 573 (2026).

Editor’s Note: Although the Court did not decide any ultimate question about the reliability or admissibility of fingerprint evidence, anyone interested in latent print identification is strongly encouraged to monitor the proceedings this opinion sets in motion. Because the Court retained jurisdiction after ordering development of a full expert record, the Special Adjudicator’s forthcoming findings and the Court’s ultimate ruling could reshape how fingerprint evidence is admitted and presented in New Jersey, notwithstanding its acceptance there for more than a century.  

Subscribe today

Already a subscriber? Login