Background: The World Health Organization (WHO) developed guidelines for psychosocially assisted pharmacological treatment of opioid dependence applying GRADE methodology. Objectives: To analyse how many recommendations are supported by Cochrane reviews (CRs), relative quality of evidence, and reasons for low quality. Methods: Clinical recommendations were extracted from the draft guidelines. Results: A total of 15 recommendations. Two recommendations for ‘choice of treatment’: 2/2 CRs evidence summarised in 3 GRADE profiles; critical outcomes: retention in treatment: high/moderate quality 3/3; use of opiate high quality 2/3; relapsed at follow up low quality 1/3. Nine recommendations for ‘maintenance’: 4/9 CRs evidence summarised in 7 GRADE profiles; critical outcomes: retention in treatment: high, moderate quality 7/7 profiles; use of opiate: high/moderate quality 3/5; mortality very low 2/2; retention at the end of follow up high 1/1; abstinence at follow up low 2/2. Four recommendations for ‘detoxification’: 4/4 CRs evidence summarised in 6 GRADE profiles; 5 critical outcomes; completion of treatment: moderate quality 5/6; side effects high/moderate quality 2/4; relapsed at follow up moderate 2/4; severity and duration of withdrawal high 1/3; use of opioid moderate 1/1. Overall, GRADE profiles considered a total of 74 pairs of outcome and relative treatment comparisons; for 31 no evidence available; 13/43 quality was low or very low; main reason was imprecise or sparse data. Conclusions: The GRADE process proved a useful way of assessing quality of evidence in this area. Good quality evidence for choosing between maintenance and detox. For maintenance, a few recommendations based on CRs, good quality only for retention, little evidence on other outcomes. No evidence on induction doses of substitution treatment and doses of buprenorphine. For detox, good quality of evidence for completion of treatment and side effects. Very little evidence, in general on long term effects, both positive and negative. CRs in this area can inform clinical guidelines for main recommendations and a few critical outcomes. Future primary research should consider a broader range of outcomes and include sizable number of patients.