OBJECTIVES: Buprenorphine/naloxone combination (BUP/NAL) is recommended in the treatment of opioid dependence. Clinical guidelines do not specify the minimum duration of treatment required to achieve long-term remission. This study evaluated the impact of treatment persistence on healthcare charges. METHODS: Study was conducted on a US insurance claims database. It included patients initiating treatment with BUP/NAL claim between November 2006 and December 2011, not previously treated with buprenorphine, with at least one repeat claim after 30 days. Discontinuation was defined as absence of BUP/NAL claim for 90 days. Healthcare charges over 12 months were compared between persistent and non-persistent patients, adjusting on baseline characteristics (demographics, comorbidities, treatment, and resource utilization before index date). RESULTS: Of 19,008 patients with an incident claim of BUP/NAL, 35.7% appeared to be short-term users and were excluded. Among the remaining 12,231 patients, the average duration of follow-up was 12.9 months, and 2846 were followed for at least two years. The probability of continuing treatment over 24 months was 40.9%. Patients under 25 years old, with a diagnosis of hepatitis or soft tissue infection were more likely to discontinue. Patients treated for at least 12 months had lower mean total charges compared to non-persistent patients ($22,912 vs. $31,687; p<0.0001), adjusting on baseline characteristics. Among non-persistent patients, total charges per quarter reached a maximum during the first trimester following discontinuation (+91% compared to period from 6 to 4 months before discontinuation, p<0.0001), and were also significantly higher in the second trimester after discontinuation (+52%, p=0.0003), compared with before discontinuation. Main drivers of excess charges were hospitalization and outpatient visits. Majority of long-term users of BUP/NAL discontinued treatment before 24 months. CONCLUSIONS: Non-persistence was associated with higher charges and evidence was consistent with a causal relationship between discontinuation and increased charges. Treatment persistence improvement may lead to cost savings.