This ecological study examines the association of the poverty status of urban communities in New York City with their mortality rates of accidental drug overdoses. Mean annual age-adjusted rates of drug overdoses involving cocaine, opiates, or both (n = 1,684) were calculated for each of 59 residential community districts in New York City for 1990-1992. A linear regression analysis was performed to test the association of the mortality rate with the poverty status of the district as measured by the proportion of the district living below the 1989 U.S. poverty line. Poverty status accounted for 69% of the variance in the drug overdose mortality rates of communities (p < .001). This study suggests that mortality rates of overdoses involving cocaine and optiates are significantly associated with the poverty status of communities in New York City.