The geographic distribution of newborn human immunodeficiency virus seroprevalence at zip code level was compared with the distribution of four sociodemographic variables. For New York City significant univariate correlations were found between HIV and low birthweight, drug use (as measured by hospital discharges), maternal education less than 12 years, and race/ethnicity. Less significant correlations were found for New York State exclusive of New York City. For New York City a model comprising low birthweight and all drug discharges was shown by multiple regression analysis to be most strongly associated with HIV status (r2 = .77). Elsewhere a model comprising race/ethnicity (percent Black, percent Hispanic) and cocaine drug discharges had the best, but less strong association (r2 = .39). However, because of intercorrelations, race/ethnicity added little when the other variables were included first. Knowledge of the geographic association between HIV seroprevalence and sociodemographic status can be useful in designing and focusing prevention efforts in areas at highest risk for future HIV/AIDS activity.