BACKGROUND: Daily oral pre-exposure prophylaxis (PrEP) is highly effective in preventing human immunodeficiency virus (HIV) infection if used adherently throughout periods of HIV risk. We estimated PrEP persistence among cohorts of persons with commercial or Medicaid insurance. METHODS: We analyzed data from the IBM MarketScan Research Database to identify persons aged 18-64 years who initiated PrEP between 2012 and 2017. We assessed PrEP persistence by calculating the time period that each person continued filling PrEP prescriptions until there was a gap in prescription fills > 30 days. We used Kaplan-Meier time-to-event methods to estimate the proportion of PrEP users who persisted with PrEP at 3, 6, and 12 months after initiation, and constructed Cox proportional hazards models to determine patient characteristics associated with nonpersistence. RESULTS: We studied 11 807 commercially insured and 647 Medicaid insured persons with PrEP prescriptions. Commercially insured patients persisted for a median time of 13.7 months (95% confidence interval [CI], 13.3-14.1), compared to 6.8 months (95% CI, 6.1-7.6) among Medicaid patients. Additionally, female sex, younger age, residence in rural location, and black race were associated with shorter persistence. After adjusting for covariates, we found that female sex (hazard ratio [HR], 1.81 [95% CI, 1.56-2.11]) and younger age (18-24 years: HR, 2.38 [95% CI, 2.11-2.69]) predicted nonpersistence. CONCLUSIONS: More than half of commercially insured persons who initiated PrEP persisted with it for 12 months, compared to a third of those with Medicaid. A better understanding of reasons for nonpersistence is important to support persistent PrEP use and to develop interventions designed for the diverse needs of at-risk populations.