Optimal methadone dosage and service profile is challenging in treatment of opioid dependence. This study explores the impact of methadone dosage on the mortality of opioid-dependent patients in methadone maintenance therapy by using a large-scale and continual supervised dosing registry information system. Database of nationwide enrolled opioid-dependent patients at methadone clinics in Taiwan during 2006-2008 was assessed. The relative risk of age, sex, marital status, HIV infection and methadone dosage were analyzed by Cox regression analysis. Among all of the 33,549 recruited patients, the crude mortality rate was 134.78/10,000 person-years, and the standardized mortality ratio was 4.68. A dose-response relationship of higher- vs. lower-dosage groups on the risk of mortality risk was observed (adjusted HR = 0.68, P = 0.016). In further sub-grouping analysis, this trend was more significant in HIV positive patients, in subgroup of patients who continuously staying in MMT, and in subgroup of patients who re-enter MMT. This dosage effect is not significantly seen in patients receiving MMT more than 365 days. Further exploration of other treatment-related factors may be important for understanding the long-term treatment outcome of opioid addiction patients.