I use a randomized experiment to test whether information can change sexual behavior among teenagers in Kenya. Providing information on the relative risk of HIV infection by partner's age led to a 28% decrease in teen pregnancy, an objective proxy for the incidence of unprotected sex. Self-reported sexual behavior data suggests substitution away from older (riskier) partners and towards protected sex with same-age partners. In contrast, the national abstinence-only HIV education curriculum had no impact on teen pregnancy. These results suggest that teenagers are responsive to risk information but their sexual behavior is more elastic on the intensive than on the extensive margin.