404 sexually active, unmarried students aged 17 through 21 from Spain, Quebec, and France were surveyed between 1989 and 1991 concerning their contraceptive usage, knowledge of AIDS, and possible risk factors in their sexual behavior. Prevention of AIDS among young people is a public health priority because relatively liberal sexual attitudes, failure to use condoms, multiple partners, and obliviousness to danger place many young people at risk. 24.6% of the students were 17, 60.5% were 18 or 19, and 14.9% were 20 or 21 years old. 62.1% of respondents from Quebec stated they always used contraception, compared to 37.6% from France and 39.3% from Spain. The proportions reporting usage of specific methods over the past year in France, Quebec, and Spain, respectively, were 21.5%, 15.6%, and 14.0% for rhythm; 35.0%, 29.4%, and 40.0% for withdrawal; 54.1%, 37.5%, and 69.6% for condom; 5.0%, 7.1%, and 5.5% for vaginal spermicides; 8.1%, 1.1%, and 9.3% for diaphragm, IUD, and cervical cap; and 62.1%, 83.5%, and 14.5% for oral contraceptives (OCs). Age at first intercourse was the only personal characteristic studied that was associated with choice of method. Respondents from Spain expressed a significantly greater fear of contracting AIDS than did those from Quebec, while the difference between Spanish and French respondents was not statistically significant. But most respondents in all countries considered their actual risk to be small. A coefficient of hazardous sexual activity was developed to assess the risk behavior of each respondent over the past six months. The measure was based on assigned values of infectivity of various practices. The total score for each respondent varied from 1 to 164, with 164 representing maximum risk. The average coefficients of hazardous sexual activity for males and females, respectively, were 64.4 and 42.2 for French, 59.1 and 69.3 for Canadian, and 32.1 and 32.7 for Spanish respondents. The coefficients were significantly lower in Spain than in the other two countries for almost all factors studied. They were usually higher in Canada than in France. Multiple regression analysis indicated that the coefficient was associated with different predictors in each of the three countries. Interventions intended to prevent AIDS should therefore take into account general social factors as well as norms of sexual behavior in each specific national context.