OBJECTIVES To examine pharmacy refill patterns of etanercept (ETA) and adalimumab (ADA) in the treatment of rheumatoid arthritis (RA) in a managed care population. METHODS Medical and pharmacy claims (January 1, 2000 – December 31, 2006) from a large managed care database were evaluated. Claims for all patients aged ³ 18 years meeting the following criteria were included: ³ two diagnosis codes for RA, no pharmacy or medical history of any biologic use for 6 months prior to anti-TNF agent index date, anti-TNF agent index date occurring on or after the first RA diagnosis date, and ³ 365 persistence days. Patients were excluded if they had a diagnosis of ankylosing spondylitis, psoriatic arthritis, psoriasis, Crohn's disease, or ulcerative colitis at anytime. Refill patterns were examined by calculating the mean time (days) between each pharmacy refill using NDC codes (actual refill days) compared to the mean days supplied on the claims (recommended refill days). Results were reported for the first year following anti-TNF agent initiation. RESULTS A total of 1239 RA patients newly starting an anti-TNF agent were included (ETA=902, ADA=337). ETA patients were slightly younger than ADA patients (ETA=48.8 years, ADA=49.2 years, p <0.0001). There was no significant gender difference between the two groups (ETA=77% female, ADA=75% female, p=0.29). Mean recommended days supplied were 32 and 34 days for ETA and ADA, respectively. Mean days between actual ETA pharmacy refills were longer than recommended for 30% of the refill periods. Mean days between actual ADA pharmacy refills were longer than recommended for 28% of the refill periods. CONCLUSIONS Approximately one-third of the actual pharmacy fills for ETA and ADA had a longer time to patient refill compared to the recommended days supply, which may indicate noncompliance.