Background: Buprenorphine/naloxone (BUP/NX), an effective treatment for opioid dependence, has been implicated in hepatic toxicity. However, as persons taking BUP/NX have multiple hepatic risk factors, comparative data are needed to quantify the risk of hepatoxicity with BUP/NX. Methods: We compared rates of alanine aminotransferase (ALT) elevation. ≥. grade 3 (ALT. ≥. 5.1 times the upper limit of normal) and graded bilirubin elevations in HIV-negative opioid injectors randomized to long-term (52 weeks) or short-term (18 days) medication assisted treatment (LT-MAT and ST-MAT, respectively) with BUP/NX in a multisite trial conducted in China and Thailand. ALT and bilirubin were measured at baseline, 12, 26, 40 and 52 weeks, times temporally remote from BUP/NX exposure in the ST-MAT participants. Results: Among1036 subjects with at least one laboratory follow-up measurement, 76 (7%) participants experienced ALT elevation. ≥. grade 3. In an intent-to-treat analysis, the risk of ALT events was similar in participants randomized to LT-MAT compared with ST-MAT (adjusted hazard ratio 1.25, 95% confidence interval 0.79 to 1.98). This finding was supported by an as-treated analysis, in which actual exposure to BUP/NX was considered. Hepatitis C seroconversion during follow-up was strongly associated with ALT events. Bilirubin elevations. ≥. grade 2 occurred in 2% of subjects, with no significant difference between arms. Conclusions: Over 52-week follow-up, the risk of hepatotoxicity was similar in opioid injectors receiving brief and prolonged treatment with BUP/NX. These data suggest that most hepatotoxic events observed during treatment with BUP/NX are due to other factors. © 2014 Elsevier Ireland Ltd.