. Patients aged ‡18 years with one or more neurologist visit with an epilepsy diagnosis and two or more pharmacy claims for AEDs were included. Medication possession ratio (MPR) was used to evaluate AED adherence with MPR ‡ 0.80 considered adherent and <0.80 considered nonadherent. The association of nonadherence with utilization outcomes [hospitalizations, inpatient days, emergency department (ED), and outpatient visits] was assessed with univariate and multivariate Poisson regressions. Quarterly per-patient inpatient, outpatient, ED, and pharmacy costs were calculated across nonadherent and adherent quarters for the younger than 65 population (under-65) and cost differences were computed. Adjusted incremental costs of nonadherence were estimated with multivariate Tobit regression models. Results: A total of 33,658 patients were included (28,470 under-65), together contributing 388,564 treated quarters (26% nonadherent). In multivariate analyses, AED nonadherence was associated with significantly higher incidence of hospitalizations [incident rate ratio (IRR) = 1.39, 95% confidence interval (CI) = 1.37-1.41], inpatient days (IRR = 1.76, 95% CI = 1.75-1.78), and ED visits (IRR = 1.19, 95% CI = 1.18-1.21). Nonadherence was associated with cost increases related to serious outcomes, including inpatient ($4,320 additional cost per quarter, 95% CI = $4,077-$4,564) and ED services ($303 additional cost per quarter, 95% CI = $273-$334), but lower costs for outpatient and pharmacy services, likely because of nonadherent behavior. Discussion: Nonadherence to AEDs appears to be associated with serious outcomes, as evidenced by increased utilization and costs of inpatient and ED services.