Simultaneous activity of agonist and antagonistic muscles during a task is known as cocontraction. The primary aim of the present study was to use a cocontraction index (CI) to quantify differences in EMG activity between a group of CP and control children at two different walking speeds. The secondary aim was to compare the amount of time the muscles were activated ("on" thresholds) between the groups. Seventeen subjects volunteered for the study. One group consisted of 9 (7M, 2F) children with CP (age 12.7 +/- 2.8 years, mean +/- SD). The second group consisted of 8 able-bodied controls (7M, 1F). The discontinuous submaximal treadmill walking protocol had two 4min stages at 0% gradient. Speeds selected were 3 km.h-1 and 90% of the pre-determined fastest walking speed (FWS). Two sites of CI were measured from the EMGs of tibialis anterior and soleus (leg) and vastus lateralis and hamstrings (thigh). Significantly (p < 0.05) higher CI values were noted for the CP subjects compared to the controls, irrespective of speed or cocontraction site and there was a significant (p < 0.05) increase in CI values with increased walking speed for both CP and control subjects. Phasic analyses for 5% max EMG and 10% max EMG "on" thresholds demonstrated significant (p < 0.05) main effects for group (CP subjects had a longer time period of muscle activation than controls) and speed (muscles were active longer at 90% FWS than 3 km.h-1). The precise mechanisms by which cocontraction contributes toward abnormal gait and wasted mechanical energy require further research incorporating both electromyographic and kinematic analysis.