A total of 128 undergraduates (56 percent females) read one of six scenarios that depicted the trial of a man accused of sexually assaulting a young girl. In the scenarios, the alleged victim either reported the assault immediately (child witness) or waited 20 years to report it (adult witness). In the adult witness condition, the woman's memory for the event had either been repressed until recently or had always been available, and expert testimony was offered on behalf of the defense, the prosecution, both, or neither. Regression analyses revealed that women perceived the accuser's testimony as more believable and the defendant's testimony as less believable than did the male participants. Similarly, the belief in the accuser's testimony decreased and the belief in the defendant's testimony increased when the accuser was an adult rather than a child, and when the defense offered expert testimony rather than not doing so. In addition, guilty verdicts were associated with higher levels of accuser believability, lower levels of defendant believability, and testimony based on repressed memories in contrast to testimony based on memories that were never repressed. Some of the issues raised by this study are identified. 4 tables and 32 references