PURPOSE: Low rates of low birthweight (LBW) among foreign-born Latinas of low socioeconomic status have been called the "epidemiologic paradox." This study examined the extent to which the paradox can be explained by differential distribution of risk factors. PROCEDURES: The data source was the 1996-1997 New York City Birth File with 78,364 singleton births to Latinas. Ancestries included Colombians, Dominicans, Ecuadorians, Mexicans, Puerto Ricans and other Hispanics. First, a logistic regression was used to predict a LBW birth with ancestry and birthplace as the only independent variables. Demographic, medical and behavioral risks were added in subsequent regression models. FINDINGS: The LBW rate for the sample was 6.8%, with significant differences between birthplace subgroups and among ancestries. Puerto Ricans had the highest LBW rates, 9.1% for the mainland-born and 9.2% for the island-born. In separate regressions for six ancestry groups, birthplace was a significant predictor of LBW only among Mexicans and other Hispanics. CONCLUSION: In this population-based study of Latina women in New York City, the positive birth outcomes of foreign-born women are largely due to their more favorable distribution of behavioral risk factors. The "epidemiologic paradox" does not account for the LBW rates among Puerto Ricans in New York City, a high percentage of whom are mainland-born (73.4%). Compared to other Latinas, Puerto Rican women are likely to have experienced far more years of acculturation, which can result in negative health behaviors.