There has been a long‐standing debate as to whether sex or human immunodeficiency virus (HIV) education actually influences the way young people behave. To the extent these programs work, they represent a potential mechanism policy‐makers might use to reduce risky behavior among youths. This paper uses data from the 2009 Youth Risk Behavior Survey to examine if students who have received school‐based HIV instruction behave differently than those who have not. To address potentially endogenous exposure to HIV education, this paper considers a propensity score matching approach. Findings from the propensity score analysis suggest that standard ordinary least squares results are biased. Despite this, there remains some evidence that exposure to HIV education decreases risky sexual activity. Among male students, HIV education is also negatively related to the rate of using needles to inject illegal drugs into the body. The needle use results are robust to a sensitivity analysis, while the results for sexual behaviors are not. (JEL H75, I18, I28, K32)