OBJECTIVE The influence of stretch of the gastrocnemiussoleus muscle on the stretch reflex activity was studied, by varying the ankle angle in steps from 10 degrees of plantarflexion (PF) to 5 degrees of dorsiflexion (DF). METHODS Nonrandomized control trial. METHODS Department of Rehabilitation Medicine of a university medical center. METHODS Sixteen subjects with and 16 subjects without spasticity. METHODS The passive elastic stiffness and active reflex response, expressed by the total and elastic path lengths, were determined at each ankle angle as a sinusoidal displacement of 5 degrees was applied to the joint at frequencies from 3 to 12 Hz. RESULTS The elastic stiffness showed no difference between the spastic and normal subjects for all ankle angles (p > .05). The elastic stiffness increased linearly similarly in both groups when the ankle was dorsiflexed. The reflex response was significantly greater in the spastic group for all positions (p < or = .01). The total and elastic path lengths showed a linear increase in both groups when the ankle angle was varied from PF to DF. The spastic group, however, had a significantly faster increase (p < or = .005). Between-group comparison showed a significant quadratic trend in the elastic path length for the spastic group (p < or = .05), with a maximum at 2.5 degrees of DF. CONCLUSIONS This study showed that the stretch reflex activity varies with the ankle position. This must be considered when performing spasticity tests subsequent to an intervention that has changed the available range of motion and when comparing subjects measured at different ankle positions.