Study participants were 573 undergraduates at Florida International University. They ranged in age from 18 to over 55 years old, with 67 percent being women. Fifty-nine percent were Hispanic, 23 percent White, 10 percent African-American, and 8 percent "other." The study used simulated trial transcripts to control the context of the event and the evidence across three types of cases (robbery, sexual assault in which the perpetrator was an acquaintance, and sexual assault in which the perpetrator was a stranger). The three simulated trials had female victims who were 6 or 13 years old, thus resulting in two transcripts of the same trial but with victims of different ages. The participants were tested in groups and were randomly assigned to read one of the six trials. Each participant received a survey that solicited demographic information, and attitudinal scales were also administered. This was followed by a reading of the trial transcript and a questionnaire that assessed the criterion measures. Participants rated the alleged victim's credibility, honesty, memory, likability, and confidence. Defendants were also rated on these measures. Global credibility was calculated for each witness. Participants who found the defendant guilty were asked to indicate the sentence they preferred. The results show that the defendant was more likely to be found guilty in the sexual assault cases than in the robbery case. The child witness was perceived to be more credible, honest, and to have a better memory in the sexual assault cases compared to the robbery case. Perceptions of memory and honesty predicted the jurors verdict and punishment. The child's age did not affect her credibility or the case verdict. Women jurors, compared to men, perceived the child to be more credible. Future research should continue to compare jurors' perceptions of children's credibility in different types of cases at different ages, using varied stimulus materials. 3 tables and 37 references