We hypothesized that neighborhood disadvantage might function as a determinant of "exposure opportunity', an intermediate step on a path toward starting to use drugs illicitly. Testing this hypothesis, we analyzed self-report data gathered in 1992 by means of confidential interviews with 1416 urban-dwelling middle-school participants in a longitudinal field study. Within this epidemiologic sample, 50 youths said that someone actively had offered them a chance to take cocaine or smoke crack; tobacco had been offered to 395 youths; alcohol to 429 youths. Using multiple logistic regression to hold constant grade, sex, minority status, and peer drug use, we found a moderately potent association between neighborhood disadvantage and exposure to cocaine: youths living in the most disadvantaged neighborhoods (highest tertile) were an estimated 5.6 times more likely to have been offered cocaine, as compared to those in relatively advantaged neighborhoods (P = 0.001). By comparison, there were weaker but statistically significant associations involving tobacco exposure opportunity (odds ratio, OR = 1.7, P = 0.004) and alcohol exposure opportunity (OR = 1.9, P = 0.0005). Future research will clarify the etiologic significance of neighborhood disadvantage in pathways leading toward illicit drug use.