Background: In several studies, patients with alcohol dependence treated with the opioid antagonist naltrexone have shown fewer relapses to heavy drinking than those receiving placebo. An interaction between the naltrexone effect and the type of psychological therapy has been observed. Methods: A 6-month, double-blind, placebo-controlled, parallel-group study was performed at 10 different investigation sites. After a placebo run-in period of 1 week, 118 patients were randomized into 4 treatment groups - 50 mg of naltrexone daily or placebo in combination with either cognitive behavioral therapy (CBT) or supportive therapy. The CBT was performed over nine sessions according to the manual of Project MATCH (Matching Alcoholism Treatments to Client Heterogeneity). The supportive therapy was defined as "the treatment as usual." Alcohol consumption, craving, carbohydrate-deficient transferrin, medication compliance by tablet count, and adverse clinical events were assessed at all visits. Other liver enzymes and psychiatric symptoms were also determined. Results: Ninety-one (77%) patients completed the study, and 92 (78%) were 80% compliant with the medication regimen. A lower percentage of heavy-drinking days was shown in the naltrexone group (p = 0.045) compared with the placebo group, as was a lower craving score (p = 0.029). These results are supported by the lower levels of liver enzyme activities (p < 0.010 for aspartate aminotransferase, alanine aminotransferase, and ?-glutamyltransferase), but not by the carbohydrate-deficient transferrin levels, in the naltrexone group. The mean time period before the first day of heavy drinking was longer for the group treated with CBT (p = 0.010), especially in combination with naltrexone (p = 0.007). Naltrexone was well tolerated, and no patients discontinued the study due to side effects. Conclusions: This study supports the effect of naltrexone in outpatient treatment of alcohol dependence and suggests that a beneficial interaction effect with CBT can be expected.