This screening trial evaluated whether amantadine hydrochloride (100 mg bid) demonstrated sufficient clinical efficacy compared to placebo to recommend development as a pharmacotherapy for cocaine dependence. Participants were randomized to amantadine (n=34) or placebo (n=35) conditions in a 16-week, placebo-controlled, double blind trial with three times per week group counseling. Amantadine-treated participants were retained significantly longer than placebo. Based on results of a joint probability index for urine drug testing results (i.e. the proportion of cocaine-metabolite free urine samples divided by the number of participants assigned to the condition), participants assigned to amantadine were found to be significantly more likely to be cocaine abstinent on the last day of 8-weeks of treatment than participants assigned to placebo. Results at the end of 16 weeks of treatment were similar. Standard measures of urine drug testing consistently favored the amantadine condition over placebo, although not at levels of statistical significance. There was no statistical significance infrequency or severity of reported adverse events by treatment condition. Participants assigned to amantadine exhibited greater reductions in global staff ratings of cocaine dependence severity from baseline to termination compared with placebo. There were no significant differences in frequency or severity of reported adverse events by treatment condition. These results provide moderate support for further study of amantadine for the treatment of cocaine dependence.