OBJECTIVE: Clinical recommendations advocate use of carbamazepine, lamotrigine, and gabapentin rather than phenobarbital and phenytoin for treating older patients with epilepsy. We describe prescribing patterns for older veterans newly diagnosed with epilepsy, determine if practice is consistent with clinical recommendations, and describe those at greatest risk of receiving these potentially inappropriate antiepileptic drugs (AEDs). METHODS: Retrospective national inpatient, outpatient, and pharmacy data from the Veterans Health Administration (VA), were used to identify veterans > 64 years with an epilepsy diagnosis during fiscal year 1999 (FY99) who also received AEDs from the VA in FY99. Patients who were seen in the VA during FY97-98 with no previous diagnosis of epilepsy were selected. We identified patients' AED regimen for FY99, demographic characteristics, neurology consultations, and disease severity. We used logistic regression to identify patients most likely to receive phenobarbital and phenytoin. RESULTS: Eighty-five percent received monotherapy. Ten percent of patients received regimens containing phenobarbital, 68% received regimens including phenytoin, and 25% received only recommended AEDs. Logistic regression analyses indicated that patients with more severe disease were less likely to receive phenobarbital monotherapy than other monotherapy (OR: 0.47, 95% CI 0.22-0.98) and phenobarbital combinations than other combinations (OR: 0.29, 95% CI 0.13-0.70). Patients receiving neurology consultation were less likely to receive phenytoin monotherapy than monotherapies consistent with clinical recommendations (New OR: 0.49, 95% CI 0.39-0.61). CONCLUSIONS: A surprising number of newly diagnosed veterans received phenobarbital despite its well known adverse effects. Moreover, our finding that nearly 70% receive phenytoin is not consistent with recent recommendations for epilepsy care in the elderly. Research is needed to identify why new patients receive phenobarbital. The vast use of phenytoin suggests research evaluating the impact of phenytoin and other AEDs on the elderly in actual practice may be needed to facilitate adoption of recent clinical recommendations.