ObjectivesTo determine predictors for sexual risk behavior among immigrant men in Amsterdam.MethodsFrom May 1997 to July 1998, 650 men from Surinam, the Antilles, and Ghana were recruited by convenience sampling. Multivariate logistic regression was used to find independent predictors for 1) multiple sex partners (>sexual 1 partner in the past year), 2) unprotected sex with casual partners in The Netherlands, 3) sexual activity in country of origin, and 4) unprotected sex in country of origin.Results1) Multiple sex partners were reported by 50% of respondents. Independent predictors were a recent infection with a sexually transmitted disease (STD), youth, low educational level, Afro-Surinamese origin, lack of a long-term relationship and history of commercial sex contact. 2) Unprotected sex occurred in 24% of casual partnerships. Independent predictors were a recent STD infection, low educational level, use of hard drugs, age between 20 and 29, few sexual partners, unprotected sex in country of origin, and a partner from the respondent's ethnic group. 3) During the past 5 years, 24% were sexually active in country of origin. Independent predictors were Ghanaian origin, age over 20, a relatively high income, and multiple partners. 4) Unprotected sex occurred in 39% of the partnerships in country of origin. Independent predictors were low educational level, history of STDs, lack of health insurance, Ghanaian origin, frequent visits to home country, few sexual partners, and unprotected sex with casual partners in The Netherlands.ConclusionAmong immigrant men in Holland, important target groups for AIDS prevention programs have been identified. Men who visit their country of origin should be made aware of the risk of sexual contact with local women. Furthermore, it should be taken into account that immigrant men use condoms less frequently with women of their own ethnic group than with Dutch women.