The Daubert Standard: Gatekeeping, Reliability, and Rule 702

The Daubert standard is the test federal courts use to decide whether expert testimony is reliable and relevant enough to reach a jury. It comes from the Supreme Court’s 1993 decision in Daubert v. Merrell Dow Pharmaceuticals, Inc., which put trial judges in charge of screening expert evidence for sound methodology and genuine connection to the facts of the case.¹ Roughly two-thirds of states have adopted Daubert or something close to it, making it the dominant admissibility standard in American litigation.

The Judge as Gatekeeper

Daubert gave trial judges a specific job: gatekeeper. Before a jury hears any expert testimony, the judge has to independently evaluate whether the expert’s reasoning and methods are scientifically sound and whether the testimony actually connects to the facts in dispute.¹ The screening usually happens at a pre-trial Daubert hearing, where both sides argue about the expert’s qualifications, methods, and conclusions.

The point of the gatekeeping role is to keep unreliable or speculative theories away from jurors, who generally lack the training to separate legitimate science from something that only sounds scientific. When a judge excludes an expert, the ruling often decides the case. If the only evidence supporting a claim comes from a witness the judge shuts out, the other side frequently wins on summary judgment.

The party offering the expert carries the burden of showing the testimony is admissible. The judge is not deciding whether the expert’s conclusions are correct. The question is whether the methodology is reliable and whether it has been properly applied to the case at hand.

The Four Reliability Factors

The Supreme Court laid out four considerations for judges evaluating whether an expert’s methodology is reliable. They are guidelines rather than a checklist, and a judge can weigh them differently depending on the type of expertise involved.¹

Testability. Can the theory or technique be tested, and has it actually been tested? A hypothesis that cannot be proven wrong through experimentation is not scientific in any meaningful sense.
Peer review and publication. Has the method been published in scientific journals and scrutinized by other professionals? Peer review does not guarantee correctness, but it means qualified people have had a chance to catch errors.
Known or potential error rate. How often does the technique produce incorrect results, and are there standards controlling how it is applied? A method with a high error rate or no quality controls is less trustworthy than one with documented accuracy.
General acceptance. Is the method widely accepted in the relevant scientific community? This used to be the sole test under the older Frye standard, and it still matters under Daubert, though it is no longer the only factor.

The list is explicitly non-exclusive. A judge can consider other indicators of reliability, and no single factor is automatically decisive. The Court built in that flexibility so the standard could accommodate fields ranging from DNA analysis to economic forecasting without forcing every discipline through the same narrow filter.

The Fit Requirement

Reliable methodology alone is not enough. The testimony also has to fit the case, meaning the expert’s knowledge must connect to the specific facts in dispute and help the jury understand something it could not figure out on its own. A perfectly valid scientific technique becomes irrelevant if it does not address the questions the jury needs to answer.

Judges also watch for analytical gaps, where an expert makes a logical leap between the data and the conclusion that the science does not actually support. This is where many Daubert challenges succeed in practice. The underlying method might be sound in the abstract, but the expert stretches it beyond what the data can reasonably show.

Federal Rule of Evidence 702

The formal requirements for expert testimony sit in Federal Rule of Evidence 702. The Daubert factors themselves were never written into the rule, but Rule 702 is the framework judges apply alongside them.² A qualified expert may testify if the proponent shows it is more likely than not that:

The expert’s knowledge will help the jury understand the evidence or resolve a factual dispute.
The testimony is based on sufficient facts or data.
The testimony is the product of reliable principles and methods.
The expert has reliably applied those principles and methods to the facts of the case.

All four requirements must be satisfied. Failure on any one can result in exclusion.

The 2023 Amendment

Rule 702 was amended effective December 2023 to add the phrase “more likely than not,” clarifying that the proponent’s burden is preponderance of the evidence under Rule 104(a).² The change responded to courts that had been applying a more lenient approach, letting questionable expert opinions reach the jury and treating reliability as a weight-of-the-evidence issue rather than an admissibility question. The committee notes stress that the proponent does not need to prove the expert’s conclusions are correct, only that the methodology is reliable and has been properly applied.³

Does Daubert Apply to Non-Scientific Experts

Yes. The original Daubert decision involved scientific testimony, which left open whether the gatekeeping obligation extended to experts whose knowledge came from technical skill or professional experience. The Supreme Court answered that question in 1999 in Kumho Tire Co. v. Carmichael, holding that the judge’s gatekeeping role applies to all expert testimony, not just testimony rooted in hard science.⁴ The Court pointed out that Rule 702 draws no distinction between scientific, technical, and other specialized knowledge, and that trying to sort experts into scientist and non-scientist categories would create an unworkable line.

For experience-based experts, judges adapt the reliability inquiry. Testability and error rates may not translate neatly, so courts instead look at whether the expert’s conclusions are grounded in an accepted body of learning in the field and whether the expert can explain how their experience leads to the conclusions offered.² An expert who simply says “trust me, I’ve been doing this for 30 years” without articulating the reasoning is unlikely to survive a challenge. The more subjective the field, the more carefully courts tend to scrutinize reliability.

How Daubert Changed the Older Frye Test

Before Daubert, courts screened expert evidence using the general acceptance test from Frye v. United States, a 1923 D.C. Circuit case. Under Frye, a judge asked one question: is the expert’s method generally accepted by other professionals in the same field? If most practitioners in the discipline endorsed the technique, the testimony came in. If they did not, it stayed out.

Frye was simple, but the simplicity cut both ways. Valid new techniques could be excluded because they were too recent to have gained widespread recognition, while outdated methods could sail through on institutional inertia. By the early 1990s, the Supreme Court concluded that the Federal Rules of Evidence demanded a more rigorous and flexible approach, and Daubert replaced Frye in federal court.

Which States Follow Daubert

Daubert is a federal standard, but roughly 33 states and the District of Columbia have adopted it or a substantially similar test for their own courts. About seven states still follow the Frye general acceptance approach, including California, New York, Pennsylvania, and Illinois. The rest apply hybrid tests that borrow from both frameworks.

The practical difference matters. In a Frye state, a novel but well-validated technique may be excluded because the relevant scientific community has not yet broadly endorsed it. In a Daubert state, the same technique could be admitted if the judge finds it testable, peer-reviewed, and reliable, even without universal acceptance. Litigants in state court need to know which standard their jurisdiction follows, because the strategy for presenting or challenging expert testimony shifts significantly depending on the test.

Appealing a Daubert Ruling

Overturning a trial judge’s decision to admit or exclude expert testimony is deliberately difficult. In General Electric Co. v. Joiner, the Supreme Court held that appellate courts review Daubert rulings under the abuse of discretion standard, the most deferential standard of review available.⁵ The same standard applies whether the trial judge let the testimony in or kept it out, and it does not change just because the ruling effectively decided the case.

To win on appeal, the losing side generally has to show the trial judge made a meaningful error in judgment, such as skipping the reliability analysis altogether, ignoring a factor that clearly mattered, or relying on an irrelevant consideration. Reversal is more likely when the trial judge did not explain the reasoning behind the ruling. A well-documented record with written findings gives the appellate court something to review; an unexplained ruling is more vulnerable. A formal Daubert hearing is not always required, but failure to hold one can be reversible error if the record is too thin for the appellate court to assess what happened.

1
Justia Law. Daubert v. Merrell Dow Pharmaceuticals, Inc. 509 U.S. 579 (1993)
2
Legal Information Institute. Federal Rules of Evidence Rule 702 – Testimony by Expert Witnesses
3
United States Courts. Federal Rules of Evidence
4
Legal Information Institute. Kumho Tire Co. v. Carmichael
5
Legal Information Institute. General Electric Co. v. Joiner