Entertainment-education (E-E) may offer an effective way to reduce risky behavior by modeling healthy behaviors. Although there is some empirical evidence to support the effectiveness of the E-E strategy, much of this research has been conducted in countries with different media landscapes than that of the United States and controlled experiments in this context are rare. Moreover, empirical tests of the relative effectiveness of E-E messages and other message formats are needed. In this study, 437 undergraduates participated in a three-wave panel experiment in which they viewed one of three programs (E-E, education, or entertainment). Safer sex intentions and behaviors were measured several days before, immediately following, and 2 weeks after exposure. Results demonstrate that effects of exposure to this E-E program vary depending on gender and past experience with sexual intercourse. In particular, females and those who had not initiated sexual intercourse showed the strongest effects. Discussion of theoretical implications and suggestions for future research are provided.