The sample was stratified into the following subgroups to represent the probability of different levels of involvement in alcoholism and alcohol-related behaviors: people in county jails, people on probation, people in alcoholism treatment centers, and people in drinking driver programs. The study was conducted between August and October 1989, and 878 respondents from five New York counties participated. Alcohol consumption was measured by the respondent's normal quantity and frequency of alcohol use. Frequency of cocaine use was measured by a five-point scale. Respondents were also asked to report the number of personal and property crimes they had committed while under the influence of alcohol. Data indicated alcohol and cocaine use tended to increase criminality more in an independent than in an interdependent manner. Effects of early alcohol and cocaine use on criminality did not appear to depend on current use, and the extent to which cocaine use affected criminality did not seem to depend on the intensity of alcohol use. Alcohol and cocaine use appeared to increase criminality independently of each other. Limitations of the research are noted. 46 references, 2 notes, and 5 tables