Background: Because of phenytoin's narrow therapeutic index and nonlinear pharmacokinetics, food-induced alterations in absorption may markedly influence drug concentrations and, in turn, safety and effectiveness. Potential food-associated differences between 100-mg Mylan (Mylan Pharmaceuticals) extended-release phenytoin sodium capsules and Parke-Davis 100-mg Dilantin Kapseals were examined. Methods: A single-dose, two-way crossover study was conducted in 24 healthy subjects to determine the effect of a high-fat meal on the pharmacokinetics of both formulations. Pharmacokinetic parameters were estimated by noncompartmental methods. The impact of switching products on steady-state phenytoin concentrations was investigated through simulation using pharmacokinetic data previously obtained from 30 epileptic patients. Results: Based on AUC(0-??), bioavailability of the Mylan product administered with food was 13% lower than that observed with Dilantin Kapseals. Simulations of substituting the Mylan product for Dilantin suggested that the 13% decrease in bioavailability would result in a median 37% decrease (range 19 to 58%) in plasma phenytoin concentrations when the drug is given with food; in 46% of patients, phenytoin concentrations would likely fall below the therapeutic range of 10 to 20 mg/L. Simulations of substituting Dilantin for the Mylan product suggested that the 15% increase in bioavailability would result in a median 102% increase (range 24 to > 150%) in plasma phenytoin concentrations, with 84% of patients having phenytoin concentrations above the therapeutic range. Conclusions: Results suggest that when taking phenytoin sodium with food, product switches may result in either side effects or loss of seizure control.