Drawing on the responses provided by a survey of state court judges (N = 400), empirical evidence is presented with respect to judges' opinions about the Daubert criteria, their utility as decision-making guidelines, the level to which judges understand their scientific meaning, and how they might apply them when evaluating the admissibility of expert evidence. Proportionate stratified random sampling was used to obtain a representative sample of state court judges. Part I of the survey was a structured telephone interview (response rate of 71%) and in Part II, respondents had an option of completing the survey by telephone or receiving a questionnaire in the mail (response rate of 81%). Survey results demonstrate that judges overwhelmingly support the "gatekeeping" role as defined by Daubert, irrespective of the admissibility standard followed in their state. However, many of the judges surveyed lacked the scientific literacy seemingly necessitated by Daubert. Judges had the most difficulty operationalizing falsifiability and error rate, with only 5% of the respondents demonstrating a clear understanding of falsifiability and only 4% demonstrating a clear understanding of error rate. Although there was little consensus about the relative importance of the guidelines, judges attributed more weight to general acceptance as an admissibility criterion. Although most judges agreed that a distinction could be made between "scientific" and "technical or otherwise specialized" knowledge, the ability to apply the Daubert guidelines appeared to have little bearing on whether specific types of expert evidence were designated as "science" or "nonscience." Moreover, judges' "bench philosophy of science" seemed to reflect the rhetoric, rather than the substance, of Daubert. Implications of these results for the evolving relationship between science and law and the ongoing debates about Frye, Daubert, Joiner, and Kumho are discussed.