Naltrexone treatment has demonstrated some advantages for special populations of heroin addicted individuals, but patients' compliance seems to be very poor, with a low adherence and low retention rate. Kappa-opioid system overdrive seems to contribute to opioid protracted abstinence syndrome, with dysphoria and psychosomatic symptoms during naltrexone treatment. The objective of this observational study was to determine the effectiveness of a functional k antagonist in improving naltrexone treatment outcome. A partial mu agonist/kappa antagonist (buprenorphine) and a mu antagonist (naltrexone) were combined during a 12 weeks protocol, theoretically leaving k antagonism as the major medication effect. Sixty patients were submitted to outpatient rapid detoxification utilizing buprenorphine and opioid antagonists. Starting on the fifth day, 30 patients (group A) received naltrexone alone. Alternatively, 30 patients (group B) received naltrexone (50mg oral dose) plus buprenorphine (4 mg sublingual) for the 12 weeks of the observational study. The endpoints of the study were: retention in treatment, negative urinalyses, changes in psychological symptoms (Symptom Checklist-90 Revised: SCL-90) and craving scores (visual analysis scale (VAS)). Thirty-four subjects (56.67%) completed the 12 weeks study. Twenty-one patients (35.0%) had all urine samples negative for opiates and cocaine. nine subjects (15.0%) had urine samples negative for cocaine and opiates for the last 4 weeks of the study. five subjects (8.3%) continued to use cocaine during the 12 weeks of the study. No significant change in pupillary diameter after buprenorphine administration was evidenced during clinical observations from baseline across the weekly measurements. Retention rates in group A (naltrexone) and group B (naltrexone + buprenorphine) at week 12 were respectively 40% (12 patients) and 73.33% (22 patients), with a significant difference in favour of group B (p= 0.018). Patients treated with naltrexone in combination with buprenorphine (B patients) showed a significantly lower rate of positive urines for morphine (4.45%) and cocaine metabolites (9.09%) than those treated with naltrexone alone (A) (25%, morphine; 33.33% cocaine) (p< 0.05; p< 0.05). Irritability, depression, tiredness, psychosomatic symptoms and craving scores decreased significantly less in Group A patients than in group B patients. The dysfunction of opioid system with kappa receptors hyper-activation provoked by heroin exposure, probably underlying dysphoric and psychosomatic symptoms during naltrexone treatment, seems to be counteracted, at least in part, by buprenorphine. The combination of buprenorphine and naltrexone may significantly improve the outcome of opioid antagonists treatment in terms of retention, negative urinalyses, and reduced dysphoria, mood symptoms and craving.