1. Naltrexone (50 mg bid, p.o.) was administered in a double-blind fashion (with placebo control) to chronic schizophrenic patients who maintained their routine neuroleptic and anxiolytic therapy. 2. Both positive and negative symptom patients who received naltrexone improved with regard to symptoms involving deterioration and social withdrawal. No significant amelioration was recorded in subjects assuming placebo relative to the same psychopathological areas. 3. Favourable results were obtained mainly from patients affected by negative symptom schizophrenia. 4. Naltrexone may have acted by direct or indirect neurochemical mechanisms related to negative symptom schizophrenia.