Background: Comorbid psychiatric disorders among opioid-dependent patients are associated with several negative outcome factors. However, outcomes of maintenance treatment have not been sufficiently established, and no evidence is available with respect to heroin-assisted treatment (HAT). Methods: For patients in the German heroin trial outcome measures were analyzed for HAT versus methadone maintenance treatment (MMT) both for patients with and without a comorbid diagnosis according to CIDI. Results: 47.2% of the sample had at least one comorbid psychiatric diagnosis, mainly neurotic, stress-related or somatoform (F4) or affective (F3) disorders. HAT had a better outcome than MMT concerning improvement of health and reduction of illicit drug use in both comorbid and non-comorbid patients, but weaker effects were found in the comorbid group. Conclusions:The better outcome of HAT also in comorbid patients suggests that psychiatric comorbidity should be an inclusion criterion for HAT. The weaker advantage of HAT may be due to pharmacological or methodological reasons.