OBJECTIVES: Although opioids and benzodiazepines are now the leading cause of prescription drug poisoning and overdose deaths in the United States, their concomitant use is increasing on a yearly basis. The objective of the study was to compare the risk of an opioid overdose requiring hospitalization in patients concomitantly using opioids and benzodiazepines to patients using opioids only. METHODS: : We identified a cohort of opioid initiators with a non-cancer pain diagnosis in a commercial claims database from 2008-14. Exposure was classified into periods of concomitant and opioid only use. For each concomitant user, we matched four control periods of opioids-only use on time since opioid initiation using an incidence density sampling approach. We estimated a propensity score on the matched sample using 86 baseline characteristics including patterns of prior opioid use. The outcome was defined as hospitalization with principal diagnosis of an opioid overdose. Using a Cox proportional hazard model with stabilized inverse probability of treatment weighting (IPTW), we calculated the hazard ratio (HR) of an opioid overdose associated with concomitant use compared to opioid only use, modelled as a time-varying covariate. RESULTS: : We created a matched cohort of 1,114,456 opioid initiators (mean [SD] age 48[11], 64% female). The unadjusted hazard ratio (HR) for an opioid overdose was 3.08 (95% CI, 2.28-4.16). After IPTW adjustment, the HR decreased to 2.36 (95% CI, 1.86-3.03), corresponding to an absolute excess risk of 12.83 (95% CI 8.18, 19.16) overdose hospitalizations per 10,000 patient years. CONCLUSIONS: : Compared to opioid use alone, concomitant use of opioids and benzodiazepines was associated with a significantly higher risk of opioid overdose requiring hospitalization; however the absolute risk difference was small. Further research should evaluate the mechanism of this association.