Aims. This paper examines the costs of medical care prior to and following initiation of alcoholism treatment as part of a study of patient matching to treatment modality. Design Longitudinal study with pre- and post-treatment initiation. Measurements. The total medical care costs for inpatient and outpatient treatment for patients participating over a span of 3 years post-treatment. Setting. Three treatment sites at two of the nine Project MATCH locations (Milwaukee, WI and Providence, RI). Participants. Two hundred and seventy-nine patients. Intervention. Patients were randomly assigned to one of three treatment modalities: a 12-session cognitive behavioral therapy (CBT), a four-session motivational enhancement therapy (MET) or a 12-session Twelve-Step facilitation (TSF) treatment over 12 weeks. Findings. Total medical care costs declined from pre- to post-treatment overall and for each modality. Matching effects independent of clinical prognosis showed that MET has potential for medical-care cost-savings. However, patients with poor prognostic characteristics (alcohol dependence, psychiatric severity and/or social network support for drinking) have better cost-savings potential with CBT and/or TSF., Conclusions. Matching variables have significant importance in increasing the potential for medical-care cost-reductions following alcoholism treatment.