We compared the effectiveness and costs of day hospital (DH) versus inpatient (INP) rehabilitation for cocaine dependence. The research subjects were 111 inner city, lower socioeconomic, primarily African-American male veterans who qualified for a diagnosis of cocaine dependence and presented no acute medical or psychiatric conditions requiring inpatient treatment. Fifty-six men were randomly assigned to 1 month of DH rehabilitation (27 hours of weekday treatment weekly), and 55 were assigned to 1-month INP rehabilitation (48 hours of scheduled treatment weekly). Treatment outcome was evaluated 7 months after admission into treatment (92% of the subjects), and a cost analysis was performed. A significantly greater proportion of INP subjects (89.1%) completed treatment than did DH subjects (53.6%). Significant improvements in substance use, psychosocial functioning, and health status were found 7 months postadmission for both groups, but there was little evidence of differential improvement between groups. Urine toxicology findings were consistent with the self-report data in showing improvement from baseline, but no group differences in cocaine use. The groups did not differ significantly in post-rehabilitation aftercare participation or in relapse to additional treatment. DH treatment costs were 40% to 60% of INP treatment costs, depending upon the measure used.