OBJECTIVES: To assess differences in rates of shorter, longer, and on-time prescription refill intervals for adalimumab and etanercept in rheumatoid arthritis (RA) and to identify factors associated with rates of longer prescription refill intervals. METHODS: An event-level analysis was conducted using medical/pharmacy claims from a database of 98 managed care plans. Inclusion criteria included index tumor necrosis factor inhibitor (anti-TNF) started January 2004-December 2007, patient age ≥18, 2 RA diagnosis codes (ICD-9 code 714.xx), and 365 days of index anti-TNF treatment. Patients with selected other inflammatory conditions or evidence of anti-TNFs during 6 months prior to index date or abatacept or rituximab while on index anti-TNF were excluded. Prescription refill intervals were assessed by comparing the observed time between dates of prescription claims with recommended dosing frequency in prescribing information (i.e., adalimumab=14 days, etanercept=7 days). Observed refills within seven days of the recommended dosing frequency were considered “on time”; intervals greater than seven days “longer”; and less than seven days “shorter”. Multivariable logistic regression analyses were performed to examine determinants of longer refill intervals. RESULTS: There were 26,103 prescription refill events for adalimumab (N=1,279 patients), and 48,859 for etanercept (N=2,277 patients). Rates of shorter refill intervals were low: 3.7% for adalimumab and 3.5% for etanercept. Longer and on-time refill interval rates were comparable for adalimumab (27.4% and 68.9%) and etanercept (29.5% and 67.0%), respectively. Regression analyses revealed etanercept longer intervals increased with duration of use (p<0.01), female gender (p<0.01), and lack of insurance (p<0.01), but decreased with older age (p<0.01). A similar pattern emerged for adalimumab. CONCLUSIONS: More than one in four of all adalimumab and etanercept prescription refill events in RA had longer refill intervals than recommended. Further research is warranted assessing the clinical and economic consequences of delays in prescription refills, which may be suggestive of under-dosing.