Recent studies suggest that gabapentin has a neuroprotective effect in experimental models of motoneuron disease. We carried out a multicenter, randomized, controlled trial of gabapentin versus no treatment in 120 patients with type II or III spinal muscular atrophy for 12 months. We assessed maximum voluntary isometric contraction with a handheld myometer and calculated an arm megascore (summing elbow flexion, hand grip, and three-point pinch scores), and a leg megascore (summing knee flexion, knee extension, and foot extension scores). Forced vital capacity and timed tasks were also evaluated. Arm megascore improved by at least 30% in 24.6% of treated and 16.9% of untreated patients (relative risk = 1.45; 95% confidence interval = 0.71-2.97). The leg megascore improved by at least 30% in 37.7% of treated and 20.3% of untreated patients (relative risk = 1.85; 95% confidence interval = 1.02-3.37). We conclude that gabapentin produced a significant improvement in leg megascore at 6 months, which was more evident at 12 months, with a trend for improvement in arm megascore at 12 months. The treatment had no effect on forced vital capacity or timed functional tests.