This study explores the associations between pharmacist-managed antiepileptic drug therapy in hospitalized Medicare patients and diagnoses indicating the need for these drugs. It also explores the following major heath care outcomes: death rate, hospital length of stay (LOS), Medicare charges, drug charges, laboratory charges, complications, and adverse drug reactions. Data were drawn from the 1998 MedPAR and 1998 National Clinical Pharmacy Services databases. Pharmacist-managed antiepileptic drug therapy was evaluated in a study population of 9380 Medicare patients with diagnosed epilepsy or seizure disorders treated in 794 United States hospitals. This population was derived from the 38,311 hospitalized Medicare patients with epilepsy or seizure disorders (MedPAR). In hospitals without pharmacist-managed antiepileptic drug therapy, death rates were 120.61% higher, with 374 excess deaths (chi(2)=5.983, df=1, p=0.014, odds ratio [OR]=1.553, 95% confidence interval [CI] 1.102-2.189). Hospital LOS was 14.68% higher, with 8069 patient-days (Mann-Whitney U test [U]=3833132, p=0.0009); total Medicare charges were 11.19% higher, with 14,372,550 dollars in excess total charges (U=3644199, p=0.0003); per-patient drug charges were $115 +/- $92 higher (p=NS); laboratory charges were 32.24% higher, with 5,664,970 dollars in excess charges; and aspiration pneumonia rate was 54.61% higher (chi(2)=5.848, df=1, p=0.015, OR=1.233, 95% CI 1.081-1.901). Although the frequencies of other complications and adverse effects were higher, these differences were not statistically significant compared with hospitals with pharmacist-managed antiepileptic drug therapy. Clinical and economic outcomes were improved among hospitalized Medicare patients whose antiepileptic drug therapy was managed by pharmacists.