OBJECTIVES: This study sought to explore the relationships between pain etiologies and the probability of long term opioid use among persons newly prescribed an opioid. METHODS:  We identified cancer-free persons with new opioid use episodes who were greater than 13 years of age without a prior history of substance abuse from a nationally representative database of commercially insured population from 2006-2015. We categorized individuals into 11 mutually exclusive pain etiology categories based on their medical claims: Trauma and Surgery, Trauma, Surgery, Burn, Delivery, Dental Procedure, Chronic pain, Non-chronic pain, Inpatient stay, Emergency department visit. Chronic pain was assessed in prior six months and all others were assessed in the week prior to receiving an opioid. Patients were followed until opioid discontinuation (a gap of at least 180 days without opioid use), eligibility loss or study end (September 2015). Cox proportional Hazards models were used to model the time to opioid discontinuation controlling for potential confounders. RESULTS:  We identified a total of 1,386,132 patients with median time to discontinuation of 5 days. Chronic pain diagnoses (25.58%) followed by non-chronic pain (24.87%) and trauma (9.89%) were the most frequent etiologies for an opioid. Persons with chronic pain had the highest probability of continued opioid use at one year (8.91%) and dental procedures had the lowest (0.92%). Relative to surgery patients, those with a chronic pain diagnosis had the lowest likelihood of discontinuation (HR=0.73;p<0.001) followed by patients who had an inpatient stay (HR=0.76; p<0.001) or surgery and trauma (HR=0.77; p<0.001). CONCLUSIONS:  Patients treated with opioids for chronic non cancer pain are the most likely to use opioid over long periods. Persons with surgery, trauma, non-trauma ED use, and deliveries prior to an opioid are infrequently prescribed opioids for long durations suggesting that managing acute pain with opioids seldom develops into chronic opioid use.