Gabapentin, an analog of gamma-aminobutyric acid, exhibits anticonvulsant properties in both animal models and humans. Gabapentin pharmacokinetics was studied in laboratory animals using HPLC and radiometry. Oral bioavailability was 40% in monkeys administered 25 mg/kg, 79% in mice and rats receiving 50 mg/kg, and 80% in dogs administered 50 mg/kg. Binding to plasma proteins was < 3%. Maximum blood or plasma concentrations generally occurred within 2 hr of an oral dose. In rats and monkeys, increases in maximum plasma concentrations and/or areas under the curve were less than dose-proportional following oral administration, most likely because of saturable absorption. However, intravenous pharmacokinetics in rats were linear over the dosage range of 4-500 mg/kg. Mean intravenous elimination half-life was 1.7 hr in rats, 2.9 hr (14C only) in dogs, and 3.0 hr in monkeys. In rats and dogs, repeated administration did not alter gabapentin or 14C pharmacokinetics. Additionally, gabapentin did not induce hepatic cytochrome P450 monooxygenases in rats. There were no age- (rats only) or gender-associated changes in pharmacokinetic parameters. [14C]Gabapentin was extensively distributed to tissues. In the dog, gabapentin was metabolized to N-methylgabapentin (approximately 34% of dose); whereas metabolism in mouse, rat, and monkey was minimal (< 5%). The principal route of excretion was via urine. In summary, as an antiepileptic drug, gabapentin exhibited desirable pharmacokinetic properties, such as linear elimination kinetics, not highly bound to plasma proteins, not extensively metabolized, and not an inducer of hepatic cytochrome P450.