In a randomized double-blind cross-over study humans with chronic spinal cord injury received ganglioside GM-1 or placebo for 2 months. GM-1, administered intravenously at a dose of 100 mg, 6 days a week, resulted in a statistically significant improvement of motor scores (P < 0.05), whether administered before or after 2 months of placebo. There was no placebo effect on motor scores. Subjects who received GM-1 before placebo maintained their improvement during the placebo phase. Subjects who received GM-I ambulated with a reciprocal gait, using orthotics, for longer distances and at a faster rate whether the drug was administered before or after placebo. These results constitute the first finding that any chemical substance improves locomotion in human chronic spinal cord injury.