The effectiveness of interventions that directly inform people of the risks posed by their own behavior is frequently undermined by people's ability to defend themselves against unwanted information. In response to this difficulty, an alternative approach to shaping people's perceptions of personal vulnerability in which people remained unaware of the intervention was tested. As part of a survey on sexual issues, college undergraduates reported the total number of sexual partners they had had on a response scale that systematically framed their sexual behavior as being either above average or below average. Participants in the 2 conditions reported the same absolute number of sexual partners. However, those who received a response scale that implied they had had more partners than the average student reported greater concern about their sexual behavior than did those who received a scale that implied they had had fewer partners than average. The implications of this approach for influencing perceptions of personal risk are discussed.