Federal Rule of Evidence 702 was amended in 2000 to provide that expert testimony is admissible only if:(1) the testimony is based upon sufficient facts or data, (2) the testimony is the product of reliable principles and methods, and (3) the witness has applied the principles and methods reliably to the facts of the case. This rule is both a codification, and to some extent an elaboration, on the so-called "Daubert trilogy" of Supreme Court cases. It's an elaboration because, for example, the Joiner decision stated that courts MAY exclude testimony if an expert witness has taken a potentially reliable principle or methodology, but not applied it reliably to the facts at hand; Rule 702 states that the court MUST do so.
Nevertheless, many federal district courts are simply ignoring or merely paying lip service to Rule 702, and relying on pre-2000 circuit court precedent. A particularly egregious example is Riley v. Target Corp., 2006 WL 1028773, slip op. (E.D. Ark. Apr. 13, 2006). No need to bore readers with the facts of the case. Target challenged the plaintiffs' physician's "differential diagnosis" (which was actually a differential etiology) under Rule 702.
The court found that the methodology of differential diagnosis is a generally reliable one, and then added that any weaknesses in the expert's opinions go to the weight of his testimony, not its admissibility. The court then cited a 1995 (!) circuit court case for the proposition that "[f]aults in an expert's use of differential etiology as a methodology or lack of textual authority for his opinion go to the weight, not the admissibility, of his testimony".
Note that the court not only relied on a pre-2000 precedent that directly contradicts the text of Rule 702, but relied on a pre-Joiner (1997) case! [UPDATE: Even worse, I just noticed, the court quoted a 1995 case for the proposition thhat "[o]nly if an expert's opinion is 'so fundamentally unsupported that it can offer no assistance to the jury' must such testimony be excluded." This was clearly wrong as of 1995, much less now, and thus not surprisingly that case itself, as the court noted, was quoting a 1988, pre-Daubert case! This judge has apparently slept through the last thirteen years of expert evidence jurisprudence.] This case is one of many examples of a court relying on stray dicta from cases that are no longer good law, and paying mere lip service, at best, to the text of Rule 702.
So note to federal district court judges, and their clerks: you shouldn't be relying on any precedent on expert evidence from before 2000, because those precedents relied on a different version of Rule 702, not to mention that many of them were decided before Kumho Tire (1999) and Joiner (which are themselves less strict than amended Rule 702). If you feel the need to do so anyway, the holding needs to be checked against the text of amended Rule 702. If the precedent conflicts with the rule, the rule quite obviously trumps the precedent.
UPDATE: There's nothing inherently wrong with citing pre-2000 precedents, so long as they are consistent with amended Rule 702. But many courts seem unable and unwilling to distinguish precedents that are consistent with Rule 702 and ones that are not. Moreover, the starting point for analysis should be the text of Rule 702, not a circuit court's interpretation of old 702 from 1998. I should add that courts should be cautious about relying even on post-2000 precedents, because many of these themselves ignore the text of 702 and instead cite pre-2000 precedents.