Low back pain is a leading cause of disability globally. It is a common reason for presentation to the emergency department where opioids are commonly prescribed. This is a retrospective cohort study of opioid-naive adults with low back pain presenting to 1 of 4 emergency departments in Nova Scotia. We use routinely collected administrative clinical and drug-use data (July 2010-November 2017) to investigate the prevalence of prolonged opioid use and associated individual and prescription characteristics. In total, 23,559 eligible individuals presented with nonspecific low back pain, with 84.4% being opioid-naive. Our study population included 4023 opioid-naive individuals who filled a new opioid prescription within 7 days after their index emergency department visit (24.4%). The prevalence of prolonged opioid use after a new opioid prescription for low back pain (filling an opioid prescription 8-90 days after the emergency department visit and filling a subsequent prescription ±30 days of 6 months) was 4.6% (185 individuals). Older age and female sex were associated with clinically important increased odds of prolonged opioid use. First prescription average >90 morphine milligram equivalents/day (odds ratio 1.6, 95% confidence interval 1.0-2.6) and greater than 7-day supply (1.9, 1.1-3.1) were associated with prolonged opioid use in adjusted models. We found evidence of declining opioid prescriptions over the study period, but that 24.3% of first opioid prescriptions in 2016 would not have aligned with current guideline recommendations. Our study provides evidence to support a cautious approach to prescribing in opioid-naive populations.