OBJECTIVES: Biological therapies are an expensive but important advance in the management of RA. The potential therapeutic benefits of biologics demonstrated in clinical trials may be undermined by poor adherence and early discontinuation of treatment (i.e., non-persistence) in clinical practice.  This study examined adherence, discontinuation, and switching of Rheumatoid Arthritis (RA) biologics over a one year period following initiation of the biologic treatment in Medicaid patients with RA.  METHODS: The study sample consisted of Medicaid patients with RA in California, Florida and New York who had newly initiated etanercept (n=1359), anakinra (n=267), or infliximab (n=1012) between January 1, 2000 and December 31, 2002.  Adherence (proportion of days covered (PDC) ≥ 0.80), discontinuation (90-day continuous gap), and switching (initiation of second biologic within 90 days of discontinuation date of index biologic) were measured during the 12-months post-index biologic initiation.  Sensitivity analyses were conducted by varying the thresholds to define these measures.  Logistic regressions examined the factors associated with RA biologic adherence and discontinuation. RESULTS: Anakinra users had the lowest mean PDC (0.36) and percent adherent patients (10.5%) followed by etanercept users (mean PDC:0.57; % adherent:32%) and infliximab users (mean PDC:0.64; % adherent:43%).  All three groups had high discontinuation rates (41% etanercept, 76% anakinra, and 41% infliximab).  Few patients who discontinued the index biologic switched to another biologic (0.2% to 9%).  Logistic regressions found that patients in Florida had lower odds of being adherent and higher odds of discontinuing their index biologic than patients in California. Consistent with descriptive results, Anakinra users had lower odds and infliximab users had higher odds of being adherent than etanercept users.  Anakinra users had higher odds of discontinuation than etanercept users. CONCLUSIONS: This study highlights poor adherence and premature discontinuation without concurrent switching of RA biologics that should raise concern for clinicians as well as payers.