OBJECTIVES: Oral anticoagulation (OAC) is the mainstay of stroke prevention in NVAF. With the arrival of NOACs as an alternative to VKA, there is interest in the utilization of these classes. This study aimed to describe treatment patterns and estimate persistence to VKA and NOACs in NVAF patients. METHODS: Retrospective cohort study (Jan 2003-Dec 2013) of patients newly diagnosed with NVAF from the Badalona Serveis Assistencials database. Change in treatment line was defined as discontinuation (absence of prescription for ≥90 days) or switch. We calculated the cumulative incidence (95% CIs) of persistence after treatment initiation by treatment line, accounting for death as a competing risk. RESULTS: Overall, 2,137 patients initiated an OAC (VKA/NOAC) during the study period, with VKA most frequently prescribed [2,127 patients (99.5%) vs. 300 patients (14.0%) initiating NOAC]. VKA was the predominant 1^st line OAC following NVAF diagnosis (2,127; 99.5%). For patients initiating 2^nd line treatments (589), half (298; 50.6%) restarted VKA after previous VKA treatment and cessation; and half (290; 49.2%) switched from VKA to a NOAC. Almost all NOAC initiations were patients previously treated with VKA (291/300 patients; 97%). Persistence to VKA was lower in 1^st line than 2^nd line patients (median time to discontinuation/switch: 156 vs. 181 days). Persistence to NOACs was not calculated for 1^st line patients due to low numbers; but for patients initiating NOAC as 2^ndline, the median time to NOAC discontinuation/switch was 241 days. CONCLUSIONS: In this study, almost all patients initiating a 1^st line OAC, initiated VKA, with essentially no first-line NOAC prescription. Unadjusted persistence was best for patients initiating a NOAC as a 2^nd line, and the shortest among patients initiating VKA as 1^st line. As more patients are treated with NOACs in Spain, future research on treatment patterns should allow adjustment of persistence comparisons between OACs.