Stiff-legged gait, ascribed to limited knee flexion during swing in spastic paresis, has previously received little detailed investigation. In this study, data from 23 patients referred for dynamic electromyographic evaluation of spastic stiff-legged gait were analyzed to identify timing of the activity of eight muscles during the gait cycle. Stride characteristics and foot switch data were also analyzed. Inappropriate activity in at least one of the quadriceps muscles during the preswing and/or initial swing phases was found in all 23 patients. Nine patients (39%) had hamstring activity during preswing. This group of 9, compared with the other 14 patients, had a significant reduction in average gait velocity and stride length (P less than 0.05) suggesting that preswing hamstring activity in stiff-legged gait may be counterproductive. No relation was found between biceps femoris (short head) activity and the amount of peak knee flexion attained in swing indicating that other factors are more important in attaining knee flexion. Delayed heel rise was observed in 21 patients (91%), which could imply insufficient calf muscle strength. Further, patients with markedly delayed heel rise achieved less peak knee flexion in swing than patients with normal or only moderately delayed heel rise (P less than 0.05). This may support the notion that adequate calf muscle strength is important in initiating knee flexion in the terminal stance/preswing phase. Results from this study provide preliminary quantitative information about stiff-legged gait that may prove useful in guiding management techniques.