We offer a new explanation for the long-run underperformance of IPO stocks using prospect theory. According to this theory, uncertain outcomes enter an investor's utility function through a nonlinear transformation of their probabilities. Small probability events are given more weight than in expected utility theory, whereas median and large probability events are given less weight. IPO stocks have more extreme positive returns; hence they are valued more in prospect theory than in expected utility theory. We test our theory with Ritter's (1991) IPO sample. Using parameter values consistent with previous experimental studies, we find investors value IPOs the same as seasoned stocks in a prospective utility setting, even though the formers' long-run average returns are much lower than the latters'.