Available data (this review includes old major articles and recent articles) show that, although results are heterogeneous, methadone maintenance treatments (MMTs) have a real efficiency not only to reduce illicit opiate abuse (50-80% of patients under MMT did not use heroin in the preceding month) but also to reduce criminality, HIV risks and mortality, and to improve social rehabilitation, without inducing other alternative substance abuse. A minority of patients (perhaps 5-20%) stay on MMT on a very long-term basis (more than 10 years). Efficiency of MMTs are rather poorly related to patients' variables, with the exception of a moderately deleterious effect of a low age at onset of opiate dependance, a precocious or high involvement in criminality and an abuse of non-opiate drugs. On the other hand, variables related to treatment play a more important role in explaining heterogeneity of results. Optimal daily dose, high quality of medical and psyche-social services, clear orientation towards social rehabilitation and treatment retention (to allow a sufficient duration of treatment) and slow detoxification regimen of well-stabilized patients are all factors contributing to better results.