Aims. To examine whether cocaine abusers differ from non-abusers in their frequency and enjoyability of engaging in various "pleasant events", in order to approximate the density of positive reinforcement experienced in their natural environment. Design. Comparisons of cocaine abusers to normative data and matched controls. Setting. An outpatient substance abuse treatment center in Burlington, Vermont, USA. Participants. Subjects included 100 individuals enrolled in outpatient treatment for cocaine abuse or dependence and 50 community volunteers without histories of drug abuse or other major psychiatric illness and matched to cocaine-dependent patients on age, sex and SES. Measurements. Diagnostic assessments were based upon clinical interviews using the DSM-III-R checklist. The primary focus of this study was the Pleasant Events Schedule (PES), a self-rated behavioral inventory of the frequency and enjoyability of engaging in "pleasant" activities. Cocaine use history, treatment outcome and other relevant variables were also assessed. Findings. Cocaine abusers reliably reported lower frequency of non-social, introverted, passive outdoor and mood-related activities than controls. These differences remained after controlling for demographic and life-style differences between groups, with the exception of mood-related activities. Perceived enjoyability of the activities did not differ across groups. Intravenous cocaine use and prior treatment for cocaine abuse predicted particularly low frequency of pleasant activities. Greater frequency of non-social activities predicted better treatment outcome. Conclusions. Drug abuse is associated with low density of certain types of non-drug reinforcement. Systematic increases in these activities may improve treatment outcome.