Outcome in 24 ambulant children with spastic diplegic cerebral palsy, in whom multilevel surgical intervention was recommended following gait analysis, is reviewed. Twelve children had surgical intervention (treatment group; eight males, four females; mean age 9 years 10 months, SD 3 years 4 months) while the other 12 did not (control group; five males, seven females; mean age 10 years 1 month, SD 2 years 11 months). All children had interval three-dimensional gait analyses (mean time between analyses: control group, 14.1 months; treatment group, 17.9 months). At follow-up the control group (mean age 11 years 9 months) showed a significant increase in minimum hip and knee flexion in stance which was not related to age, the interval between analyses, changes in the passive joint range of motion, nor changes in anthropometric measurements. The treatment group (mean age at follow-up 11 years 3 months) showed a significant improvement in minimum knee flexion and in ankle dorsiflexion in stance. Parents of nine children said their child's walking distance had increased following intervention. Of five children using posterior walkers preoperatively, two continued to use them postoperatively; two were using crutches or sticks and the remaining child walked independently. Two children who walked independently preoperatively used sticks postoperatively for community ambulation. The deterioration seen in the kinematics of the control group suggests that previous outcome studies comparing postoperative gait with preoperative gait have underestimated the immediate effects of surgery. It also raises concerns about the long-term effects of surgical intervention.