Gabapentin (GBP) is a antiepileptic drug (AED) indicated as adjunct therapy for treatment of partial seizures, with and without secondary generalization, in patients 12 and older with epilepsy. GBP (1-(aminomethyl) cyclohexaneacetic acid) is structurally related to gamma-aminobutyric acid (GABA), which readily crosses the blood-brain barrier. Radiolabeled GBP binds throughout the central nervous system in anatomic areas important in treatment of seizures. Its precise mechanism of action is unknown. An open-label, dose-ranging study of doses up to 1,800 mg produced > or =50% seizure reductions [responder rate (RR)] in 29% of patients with partial seizures. Three double-blind, placebo-controlled, parallel add-on trials at doses of 300-1,800 mg have produced RR of up to 28%, with a placebo RR of 8-10%. An active controlled, parallel group comparison of 600 mg to 2,400 mg in monotherapy conversion design showed no significant difference among the 600 mg, 1,200 mg, and 2,400 mg groups compared to a placebo group. An inpatient, active-controlled comparison of 300 mg and 3,600 mg in a parallel-design monotherapy trial showed that time to exit from the study was significantly longer for the 3,600-mg group and the completion rate significantly higher (53% vs. 17%) for patients receiving 3,600 mg/day vs. 300 mg/day of GBP. Successful double-blind, placebo-controlled trials in refractory childhood partial seizures and benign childhood epilepsy with centrotemporal spikes have been recently concluded. Absence was not successfully treated in one small double-blind trial. Open-label reports emphasize adjustments of patients to higher doses than those indicated in the package labeling. An open-label trial of GBP therapy in patients with partial seizures (n = 2,216) produced progressively greater seizure freedom rates as patients were titrated from > or =900 mg daily to > or = 1,800 mg daily (15.1% vs. 33.4%), with a similar effect on RR (18.1% vs. 44.9%). An add-on, open-label study treating partial seizures (n = 141) reported an RR of 71%, with 46% seizure-free in the last 8 weeks of treatment and doses up to 2,400 mg daily. A comparison trial of three doses of GBP to 600 mg of carbamazepine showed similar retention rates for 1,800 mg of GBP and 600 mg of CBZ. Another study reported 48% of patients experiencing 50% reduction, nine of whom had doses greater than 2,400 mg. Treatment in children has reported a 34.4% RR in 32 children with refractory partial seizures. A French open-label adjunctive trial documented a 33.9% RR; 13.4% were seizure-free during the evaluation period. Adverse experiences most commonly noted included somnolence, dizziness, and ataxia. Weight gain was sometimes reported with higher doses of GBP, and pediatric reports cite prominent behavioral changes, including hyperactivity, irritability, and agitation. GBP appears best used at doses at and potentially above those suggested in its package labeling. Although efficacy occurs at lower levels, increased GBP doses are associated with additional efficacy. Reports suggest that initiation at 2,400 mg or 3,600 mg may not be associated with increased adverse experiences. Titration to 900 or 1,200 mg on the first day of GBP therapy appear to be well tolerated.