OBJECTIVES: Treatment for opioid use disorder (OUD) imposes a substantial burden on the healthcare system. Non-adherence and misuse of buprenorphine medication assisted therapy (B-MAT) further augment costs while simultaneously reducing the benefit of treatment. This study estimated the impact of relapse on healthcare service utilization and costs.   METHODS: Administrative claims for members with a diagnosis of opioid dependence (ICD-9 304.0 and 304.7), abuse (305.5), or poisoning (965.0) were provided by Aetna (Blue Bell, PA) from Q1 2012 through Q1 2015.  Newly initiating B-MAT members were identified and examined in the 12-month period following initiation for relapse indicators, service utilization, and expenditure.  Relapse was defined as the appearance of one of four proxies: ER, IP, or detoxification claim with an OUD diagnosis, change in OUD diagnosis sub-classification from “in remission” to “continuous” or “episodic”, presence of any opioid fill in the absence of trauma or surgery, an opioid prescription claim(s) covering ≥ 30 consecutive days, and abrupt discontinuation of B-MAT indicative of improper tapering.  RESULTS: Overall, 60.4% of the sample exhibited at least one relapse event. Abrupt discontinuation of B-MAT, opioid use without surgery, and hospitalization were the most common relapse proxies observed. Compared to non-relapsing members, relapsing members exhibited significantly increased utilization for all services, except for pharmacy, and incurred 2.5 fold higher total healthcare costs (ps < 0.05). Inpatient hospitalization, which made up 36% of relapsing member’s costs, was the largest cost driver, at 27 times that of the non-relapsing group (p < 0.01). Outpatient hospital and ER utilization were also significantly elevated compared to the non-relapsing group (ps < 0.05). CONCLUSIONS: Relapse in OUD is prevalent, costly, and can jeopardize success with B-MAT. Improved population health management within the OUD population could prevent relapse and promote positive patient outcomes, which may reduce unnecessary healthcare spend.