The study determined how drug-dealing activity is geographically related to rates and places of self-reported drug use after controlling for sociodemographics and drinking behaviors. Data were obtained for the period 1991 through 1993 from the Fighting Back community evaluation, which assessed the impact of community-wide interventions designed to reduce alcohol and drug use. As part of the evaluation, a general population survey included telephone interviews with individuals from 21 communities, 20 of which were urban metropolitan areas. Responses from 16,083 individuals were analyzed at the zip code level (n=158). Analyses were conducted separately for youth and adults, using spatial regression techniques. The dependent variable was the percentage of respondents using drugs in the past year. The major independent variable, neighborhood drug availability, was measured by the percentage of nondrug users who had been approached to purchase drugs. The findings showed that for youth (ages 12 to 18), the level of drug sales in neighborhoods adjacent to and surrounding areas were the youth lived was positively associated with youths' self-reported drug use. For adults, drug sales within the neighborhoods where they lived were negatively associated with self-reported drug use, but drug sales in immediately adjacent neighborhoods were positively related to adults' self reports of drug use. The findings suggest that the areas where rates of drug users are highest are not necessarily the areas where drugs are sold. Further research should explore the location and types of places where drug users purchase drugs in relation to the places where they are more likely to use drugs. 3 tables and 16 references