This study examined whether participants were sensitive to variations in the quality of an experiment discussed by an expert witness and whether they used heuristic cues when evaluating the expert evidence. In the context of a hostile work environment case, different versions of the expert testimony varied the presence of heuristic cues (i.e., whether the expert's research was generally accepted or ecologically valid) and evidence quality (i.e., the construct validity of the expert's research). Men who heard expert testimony were more likely to find that the plaintiff's workplace was hostile than were men who did not hear the expert testimony; expert testimony did not influence women's liability judgments. Heuristic cues influenced participant evaluations of the expert testimony validity, but evidence quality did not. Cross-examination did not increase juror sensitivity to evidence quality. Implications for science in the legal system are discussed.