OBJECTIVES: To evaluate switching, discontinuation, and persistence among patients with RA receiving anti-TNFs. METHODS: This was a retrospective cohort study of RA patients, in a large managed care database, using their first anti-TNF January 1, 2005 -June 30, 2006.  Patients were required to have ≥30 months of continuous plan eligibility: 6 months prior to and 24 months following their index anti-TNF date, and no evidence of pre-index biologic use.  For patients with index adalimumab (ADA) or etanercept (ETA), a gap in days’ supply exceeding 60 days (i.e., no fills between prescription fill dates and prescription fill date + days supply + 60) indicated discontinuation. For patients with index infliximab (IFX), a gap in medical claims for infusions > 8 weeks + 60 days indicated discontinuation. Switching was defined as use of a new biologic post-index.  Persistence was defined as continuous use of index anti-TNF without discontinuation or switching.  Differences in switching, discontinuation, and persistence rates were evaluated.RESULTS: A total of 1780 patients were analyzed: ADA= 601 (33.8%); ETA= 785 (44.1%); IFX =394 (22.1%). Over three-quarters (77%) were female; mean age was 50.1 years. IFX patients were older and had a higher Charlson-Quan comorbidity score than ADA or ETA patients. Each treatment group experienced switching (ADA=13.1%, ETA=11%, IFX=13.5%). The IFX group had the lowest rate of discontinuation (37.6%) and greatest persistence (49%) compared to ADA (57.2% discontinuation, 29.6% persistence; p<0.0001 for both) or ETA (57.5% discontinuation, 31.6% persistence; p<0.0001 for both).  CONCLUSIONS: These results indicate that among RA patients receiving anti-TNFs, the portion that switch does not vary by treatment in the 2 years following start of therapy. However, differences in discontinuation and persistence were noted.  Studies are needed to investigate reasons for and impact of discontinuation and persistence on clinical and economic outcomes.