In this paper we examine evidence for two potential descriptions of juror reactions to probabilistic DNA evidence. The error-based description posits that jurors commit systematic logical or mathematical errors when they are called upon to evaluate quantitative evidence. The expectancy-based description posits that jurors use their background knowledge and beliefs in evaluating results from scientific tests. Consistent with the error-based description, participants in our study incorrectly aggregated separately presented probabilities and afforded probabilistic evidence less weight than would be expected by applying Bayesian norms. Consistent with the expectancy-based description, participants' background beliefs about the possibility of laboratory errors and intentional tampering affected the weight participants afforded a DNA match report. We discuss potential implications of these findings for the legal system and suggest directions for future research.