OBJECTIVES Access to prescription drugs may be controlled by formulary step edits. Specific criteria are sometimes set by insurers for drug coverage. Exenatide coverage may require previous prescriptions for oral antidiabetics (OAD), typically either metformin (MET), sulfonylureas (SULF) or thiazolidediones (TZD) within specific intervals or "look-back windows". These typically range from 60 to 180 days. This analysis assessed the timing of OAD claims prior to initiation of exenatide. METHODS Data were obtained from the HealthCore Integrated Research DatabaseTM. Adult patients had a claim for exenatide between May 1, 2005 and March 31, 2008, ≥12 months pre-index eligibility, and a type-2 diabetes diagnosis (N = 22,533). RESULTS A pre-index OAD claim was present for 20,228 (95%) patients. The mean (±SD) duration for any previous claim of either MET, SULF or TZD was 60±116. For MET, SULF and TZD individually, durations were 95±168, 116±202, and 135±200 days, respectively. For patients ≥ 65 years (n = 2301) the mean (±SD) duration for any previous claim for either MET, SULF or TZD was 63±114 days and for MET, SULF and TZD individually, durations were 110±174, 104±169, and 144±199 days, respectively. Of the entire cohort 6280 (28%) did not have a claim for MET, SULF or TZD within 60 days of exenatide. Of those with an OAD claim, 3975 (20%) did not have a claim for MET, SULF or TZD within 60 days. CONCLUSIONS These data represent only successfully adjudicated exenatide claims. Plans in this database may have had OAD step edits in place at some time during the study period. For plans with no OAD step edits, mean durations may have been longer and the percent of patients with fills after 60 days higher. This analysis supports careful consideration of look-back window durations. Overly restrictive intervals (i.e., 60 days) may inappropriately exclude patients from receiving exenatide.