A novel Oregon Medicaid policy guiding back pain management combined opioid restrictions with emphasis on non-opioid and non-pharmacologic therapies. To examine the effect of the policy on prescribing, health outcomes, and health service utilization. Using Medicaid enrollment, medical and prescription claims, prescription drug monitoring program, and vital statistics files, we analyzed the policy's association with selected outcomes using interrupted time series models. Adult Medicaid patients with back pain enrolled between 2014 and 2018. The Oregon Medicaid back pain policy. Opioid and non-opioid medication prescribing, procedural care, substance use and mental health conditions, and outpatient and inpatient healthcare utilization. The policy was associated with decreases in the percentage of Medicaid enrollees with back pain receiving any opioids (- 2.68 percentage points [95% CI - 3.14, - 2.23] level, - 1.01pp [95% CI - 1.1, - 0.92] slope), days of short-acting opioid use (- 0.4days [95% CI - 0.53, - 0.26] slope), receipt of more than 7days of short-acting opioids (- 2.36pp [95% CI - 2.76, - 1.95] level, - 0.91pp [95% CI - 1, - 0.83] slope), chronic opioid use (- 1.27pp [95% CI - 1.59, - 0.94] level, - 0.46 [95% CI - 0.53, - 0.39 slope), and spinal surgeries and procedures. Among secondary outcomes, we found no increase in opioid overdose and a small, statistically significant trend decrease in opioid use disorders. There were small increases in non-opioid substance use and mental health diagnoses and visits but no increase in self-harm. A state Medicaid policy emphasizing evidence-based back pain management was associated with decreases in opioid prescribing, spinal surgeries, and opioid use disorder trends, but also short-term increases in mental health encounters and an increase in non-opioid substance use disorder trends. Such policies may help reinforce evidence-based care, but must be designed with consideration of potential harms.