Background: Zonisamide is a sulfonamide antiepilepsy drug with sodium and calcium channelblocking actions. Experience in Japan and a previous European double-blind study have demonstrated its efficacy against partial-onset seizures. Methods: A randomized, double-blind, placebo-controlled trial enrolling 203 patients was conducted at 20 United States sites to assess zonisamide efficacy and dose response as adjunctive therapy for refractory partial-onset seizures. Zonisamide dosages were elevated by 100 mg/d each week. The study design allowed parallel comparisons with placebo for three dosages and a final crossover to 400 mg/d of zonisamide for all patients. The primary efficacy comparison was change in seizure frequency from a 4-week placebo baseline to weeks 8 through 12 on blinded therapy. Results: At 400 mg/d, zonisamide reduced the median frequency of all seizures by 40.5% from baseline, compared with a 9% reduction (p = 0.0009) with placebo treatment, and produced a ≥50% seizure reduction (responder rate) in 42% of patients. A dosage of 100 mg/d produced a 20.5% reduction in median seizure frequency (p = 0.038 compared with placebo) and a dosage of 200 mg/d produced a 24.7% reduction in median seizure frequency (p = 0.004 compared with placebo). Dropouts from adverse events (10%) did not differ from placebo (8.2%, NS). The only adverse event differing significantly from placebo was weight loss, though somnolence, anorexia, and ataxia were slightly more common with zonisamide treatment. Serum zonisamide concentrations rose with increasing dose. Conclusion: Zonisamide is effective and well tolerated as an adjunctive agent for refractory partial-onset seizures. The minimal effective dosage was 100 mg/d, but 400 mg/d was the most effective dosage.