Objectives: To estimate healthcare resource utilization, associated costs, and number needed to harm (NNH) from a physician's decision to prescribe extended-release (ER) non-abuse-deterrent opioids (non-ADO) as compared to ER ADOs in a chronic pain population. Design: A 12-month probabilistic simulation model was developed to estimate the reduction of misuse and/or abuse from a physician's prescribing decisions for 10,000 patients. Model inputs included probabilities for opioid misuse and/ or abuse-related events, opioid discontinuation, and switching from ADO to non- ADO. Estimated reductions in abuse associated with ADOs were obtained from positive subjective measures using human abuse liability studies. The model was run separately for commercial, Medicare, Medicaid, and Veterans Administration (VA) populations. The difference in healthcare resource utilization and associated costs (2015 USD) between the ADO and non-ADO simulations was calculated. NNH for non-ADO was also calculated. Results: Misuse and/or abuse-related events for patients prescribed ER non-ADOs ranged from 223 − 1,410 and associated costs ranged from $20 − $98 per patient for commercial and Medicare populations, respectively. Prescribing ER ADOs were associated with 87, 289, 264, and 417 fewer misuse and/or abuse − related events, saving $8, $35, $21, and $29 per patient in commercial, VA, Medicaid, and Medicare populations, respectively. NNH ranged from 185 in the commercial population to 40 in the Medicare population. Results were sensitive to decreases in the probability of misuse and/or abuse events but showed reductions. Conclusions: A physician's decision to prescribe ER ADOs could lead to large reductions in misuse and/or abuse-related events and associated costs across many patient populations.