OBJECTIVES: To compare drug acquisition costs and most commonly used doses between two GLP-1 agonists, once-weekly exenatide (QW) 2.0 mg with no dose titration, and once-daily liraglutide with a recommended starting dose of 0.6 mg and titration to 1.2 mg or 1.8 mg over 2-weeks.  METHODS: This cohort study included patients with type 2 diabetes initiating exenatide QW or liraglutide between Jan 2012 and Feb 2014, who had > 6-months of preceding coverage in U.S. healthcare claims database. During the 6 months following treatment initiation, drugs dispensed and associated pharmacy costs were compared between exenatide QW 2 mg dose and liraglutide dose subgroups, defined by maintenance dose at 60 days. RESULTS: The exenatide QW (N=3,217) and liraglutide (N=10,954) cohorts had comparable characteristics at baseline.  Due to titration recommendations, the majority of the liraglutide patients received 1.8 mg per day dose (56.2%), while 2.2% and 41.6% received 0.6 mg and 1.2 mg per day, respectively, by the end of follow-up. Comparing costs associated with exenatide QW versus specific liraglutide maintenance doses, patients on liraglutide 1.8 mg per day had higher average pharmacy costs of $3,827 (95%CI: $3,752-$3,902) compared to exenatide QW costs of $3,556 (95% CI: $3,462-$3,649), and higher average pharmacy costs associated with all antihyperglycemic drug dispensings, $2,543 (95%CI: $2,502 - $2,585), compared to  exenatide QW costs $2,345 (95%CI: $2,290 - $2,401).  CONCLUSIONS: Exenatide QW, which does not require titration, provided savings in both overall pharmacy costs and costs of antihyperglycemic drugs compared to the majority (56.2%) of liraglutide patients who required titration to the highest dose of 1.8 mg.  Lower pharmacy costs in addition to overall safety and efficacy profile, convenience of no titration, and once weekly dosing, provide a good cost-benefit value for patients treated with exenatide QW.