This study was designed to 1) determine whether isokinetic dynamometry is appropriate and safe for ambulatory patients with multiple sclerosis and 2) describe the idiosyncrasies the patients with multiple sclerosis exhibited during testing. Ten patients with multiple sclerosis and 20 healthy subjects were matched for age, sex, and weight. The isokinetic peak torque and endurance measurements of knee extensor and knee flexor muscles were recorded at selected angular velocities ranging from 0 to 275 degrees/sec and extensor:flexor muscle strength ratios were calculated. Statistical analysis (t tests) revealed that both the extensor and flexor muscle peak torque values for patients with multiple sclerosis were significantly lower than those for healthy subjects at all angular velocities (p less than .005 to p less than .001), and yet were similar in shape and slope. Mean extensor:flexor ratios for healthy subjects and for patients with multiple sclerosis were not significantly different at 70, 190, and 230 degrees/sec. Individual chart recordings identified any contractile difficulties or increased contraction time. Isokinetic testing is judged to be a safe and reliable form of objective neuromuscular testing for ambulatory patients with multiple sclerosis.