Background/objective: Few studies have examined cost of illness of epileptic partial onset seizures (POS) from the employer perspective or compared users of gabapentin and pregabalin in treatment of POS. This study compares pharmacotherapy, direct/indirect costs, and work absences of patients with POS newly started on gabapentin or pregabalin. Methods: Data from employees and dependent spouses with POS starting treatment (index date) with either gabapentin or pregabalin were analyzed. Patients were required to have at least 6 months of health plan enrollment pre- and post-index. Regression modeling compared medical and prescription costs, sick leave (SL), short-term disability (STD), workers' compensation (WC) costs and absence days during the 6-month post-index period. Persistence, adherence (proportion of days covered), impact on adherence of copay, and copay as a percent of salary were modeled. Results: Semiannual medical, drug, SL, STD, and WC costs for gabapentin vs. pregabalin cohorts were $10,306 vs. $9186, $2353 vs. $3387 (P= 0.01), $552 vs. $342, $1280 vs. $580, and $170 vs. $30, respectively. SL days (10.8 vs. 1.5, P= 0.04) and STD days (9.8 vs. 6.2) were lower in the pregabalin cohort. Persistence (median 94 vs. 70 days) and proportion with ≥80% adherence (30% vs. 15%, P= 0.049) were greater in the pregabalin cohort. Adherence decreased as copay or copay as a percent of salary increased beyond specific levels in both cohorts. Conclusion: Despite higher acquisition costs for branded pregabalin over generic gabapentin, overall direct and indirect costs trended lower for pregabalin users. Additionally, pregabalin users had significantly fewer sick leave days and significantly higher adherence rates than gabapentin users. © 2012 Elsevier B.V.