Our objective was to examine the cost of long-term residential (LTR) and outpatient drug-free (ODF) treatments for cocaine-dependent patients participating in the Drug Abuse Treatment Outcome Studies (DATOS), calculate the tangible cost of crime to society, and determine treatment benefits. Subjects were 502 cocaine-dependent patients selected from a national and naturalistic nonexperimental evaluation of community-based treatment. Financial data were available for programs from 10 US cities where the subjects received treatment between 1991 and 1993. Treatment costs were estimated from the 1992 National Drug Abuse Treatment Unit Survey (NDATUS), and tangible costs of crime were estimated from reports of illegal acts committed before, during, and after treatment. Sensitivity analyses examined results for three methods of estimating the costs of crime and cost-benefit ratios. Results showed that cocaine-dependent patients treated in both LTR and ODF programs had reductions in costs of crime from before to after treatment. LTR patients had the highest levels and costs of crime before treatment, had the greatest amount of crime cost reductions in the year after treatment, and yielded the greatest net benefits. Cost-benefit ratios for both treatment modalities provided evidence of significant returns on treatment investments for cocaine addiction.