Admissibility Standards Around the World

In 1923, the United States Court of Appeals for the DC Circuit decided Frye v United States, a case involving the then-novel polygraph. The court stated that for scientific evidence to be admissible, the technique must be sufficiently established to have gained general acceptance in the particular field in which it belongs. This formulation became the dominant American admissibility standard for most of the twentieth century and remains the law in a significant number of US states today, including California (where it is known as the Kelly-Frye test), Florida, Illinois, and New York.

Frye has identifiable virtues. It defers to the scientific community rather than asking judges to assess methodology directly. It produces relatively predictable outcomes: if a technique is established in its field, it comes in; if it is novel and contested, it stays out until the field reaches consensus. Critics argue, though, that it is too conservative (blocking genuinely valid new techniques), too permissive (admitting junk science once it has been widely adopted by practitioners), and that it conflates practitioner acceptance with empirical validation. A technique can be used in thousands of laboratories without having been rigorously tested for its error rate.

In 1993, the US Supreme Court held in Daubert v Merrell Dow Pharmaceuticals Inc that the Federal Rules of Evidence, specifically Rule 702, had superseded Frye in federal courts. Daubert identified four non-exclusive factors a judge should consider when acting as gatekeeper: whether the theory or technique can be and has been tested; whether it has been subject to peer review and publication; the known or potential rate of error and the existence of standards controlling the technique's operation; and whether it has been generally accepted. General acceptance became one factor among several rather than the sole criterion.

The Supreme Court later extended Daubert to technical and other specialised knowledge in Kumho Tire Co v Carmichael (1999), and Congress amended FRE 702 in 2000 to codify the gatekeeping requirement. A further amendment effective December 2023 tightened the rule: it now states that expert opinion testimony is admissible only if the proponent has demonstrated, by a preponderance of the evidence, that the requirements of the rule are satisfied. This shifts the burden explicitly onto the party seeking admission and was intended to reverse lower court decisions that had treated FRE 702 as merely directing the jury to decide reliability rather than the judge.

Daubert has generated a large body of jurisprudence on specific forensic techniques. Courts have admitted DNA profiling and excluded or substantially restricted bitemark comparison, hair microscopy without DNA confirmation, and some firearms toolmark opinions. The framework compels courts to engage with the science in a way Frye did not, but it has also produced inconsistent results because judges vary in their ability and willingness to scrutinise methodology.

English law has never adopted a Frye or Daubert-style admissibility test. Expert evidence is in principle admissible whenever the witness is sufficiently expert to assist the court on a matter beyond the knowledge of the judge or jury. Admissibility is governed by common-law principles of relevance and the threshold of expertise, supplemented by the Criminal Procedure Rules Part 19 (CPR 19), the Criminal Practice Directions, and the court's general discretion under section 78 of the Police and Criminal Evidence Act 1984 (PACE) to exclude prosecution evidence whose admission would have an adverse effect on the fairness of proceedings.

CPR 19 imposes a duty on the expert to help the court rather than the instructing party. The expert report must state the expert's qualifications, the literature and other material on which they rely, any range of opinion on the matter and the reasons for the expert's own view, a summary of conclusions, and a statement that the expert understands and has complied with their duty to the court. Where there is a genuine dispute about the reliability of the underlying technique, the court may order a pre-trial hearing, or the defence may apply to exclude the evidence under PACE section 78.

The Court of Appeal case R v Dlugosz [2013] EWCA Crim 2 is an example of English courts scrutinising the empirical basis of a technique without a formal admissibility test. The court held that low-template DNA evidence was admissible, but only after examining the validation data, the protocol controls, and the laboratory's compliance with its own procedures. This is functionally similar to Daubert reasoning but framed in terms of the weight and reliability of the evidence rather than a formal gateway. The Law Commission's 2011 report recommended introducing a reliability-based admissibility test in England and Wales, but that recommendation has not been enacted.

India's law of evidence was governed by the Indian Evidence Act 1872 until the Bharatiya Sakshya Adhiniyam 2023 (BSA 2023) came into force. Section 39 of the BSA 2023 (corresponding to the former section 45 of the Indian Evidence Act) provides that when the court has to form an opinion upon a point of foreign law, science, art, or on identity of handwriting or finger impressions, the opinions of persons specially skilled in such matters are relevant facts. The court then decides what weight to assign to the opinion.

India's model is weight-based rather than gateway-based. There is no formal pre-admission reliability test. A forensic scientist's opinion on, for example, DNA profiling or chemical analysis comes before the court on the same terms as any other expert opinion: the court hears the evidence and then decides how persuasive it is. The reliability of the method, the qualifications of the expert, the laboratory's accreditation status, and any cross-examination about method limitations all feed into the weight the court gives the opinion. This means an unreliable technique is not barred before trial but may be discounted or disbelieved by the judge.

The Bharatiya Nagarik Suraksha Sanhita 2023 (BNSS 2023), which replaced the Code of Criminal Procedure 1973, introduced provisions requiring electronic evidence to be accompanied by a certificate under section 63 of the BSA 2023, strengthening the authentication requirement for digital forensic evidence. The Digital Personal Data Protection Act 2023 adds a further overlay for evidence derived from personal data processing. These reforms modernise the evidentiary framework without introducing a Daubert-style admissibility gateway for science generally.

The four systems differ most sharply on three dimensions: who decides reliability, when that decision is made, and what the consequence of failure is. Under Daubert and FRE 702, the judge decides reliability before trial, and a failure means the evidence is excluded entirely. Under Frye, the judge decides at the admissibility stage but defers to the scientific community rather than conducting an independent assessment. Under English law, the judge may decide pre-trial (via a pre-admission ruling or PACE exclusion) but more often the reliability question surfaces at trial and affects weight. Under the BSA 2023, reliability is almost always a weight question decided by the fact-finder at the end of the trial.

There is a practical convergence across all four systems. Courts everywhere are asking more searching questions about forensic methodology than they did before the 2009 NAS report. Forensic scientists testifying internationally need to be prepared to explain their method's validation, its error rate if quantified, the standards they followed, and the limits of their conclusion. The legal test differs by jurisdiction; the underlying demand does not.

The post-2009 reform wave in forensic science standards was driven partly by the criminal justice systems themselves and partly by external review. The 2009 US National Academy of Sciences report Strengthening Forensic Science in the United States identified pattern-comparison disciplines as lacking adequate empirical foundations. In England, successive reviews of miscarriage of justice cases, including the Birmingham Six and the Maguire Seven, contributed to the creation of the Forensic Science Regulator role. In India, the National Forensic Sciences University has been working toward accreditation frameworks consistent with ISO 17025. The direction in each case is the same: greater transparency, documented validation, and clearer language about uncertainty.

For the working forensic scientist, the practical implication is that the formal legal test of the jurisdiction is less important than the underlying question every court is now asking. Can you describe the method you used? Has it been validated, and what is its known error rate under the conditions in which you applied it? Did you follow a documented protocol? What does your result actually mean, and what does it not mean? A scientist who can answer these questions clearly will satisfy a Daubert hearing, meet the CPR 19 duties in England, and give the Indian court the material it needs to assign appropriate weight. A scientist who cannot answer them will struggle in any jurisdiction.

See also The Daubert Standard and Its Progeny for detailed coverage of the case law applying the Daubert factors to specific forensic disciplines, and Who Is an Expert Witness for the qualification threshold that precedes any admissibility analysis.