- UNLABELLED: It has been established in able-bodied children that traditional biomechanical descriptors of gait such as stride length or stride frequency do not fully account for the differences seen in the energy cost of locomotion noted with age. Hence, measures of total body mechanical power output have been adopted to explain these differences. PURPOSE: The aim of this study was to estimate the ability of this mechanical power calculation to explain the variability in the metabolic energy cost of treadmill walking in children with spastic cerebral palsy (CP). METHODS: Thirteen subjects volunteered for the study. One group consisted of eight (6 male, 2 female) children with CP (age 12.2 +/- 2.7 yr). The second group consisted of five (4 male, 1 female) able-bodied controls (age 13.4 +/- 2.8 yr). The treadmill walking protocol consisted of one 4-min stage at 0% grade, 3 km x h(-1). Infrared markers were placed on 12 anatomical landmarks and data were collected using the OPTOTRAK motion analysis system over a 5-s time period during the last 30 s of the 4-min stage. On-line oxygen consumption VO2 measurements were obtained throughout using the Beckman Horizon Metabolic Cart. RESULTS: Relative VO2 (mL x kg(-1) x min(-1)) was significantly (P < 0.05) different between the two groups (CP: 16.6 +/- 6.5 vs control: 10.2 +/- 1.2). Simple linear regression analysis demonstrated that mechanical power measurements, incorporating transfers of energy between and within adjacent body segments, accounted for 87.2% of the total variability noted in VO2 for the children with CP, compared with only 2.4% in the able-bodied subjects. CONCLUSIONS: The results indicate that mechanical power differences explain the majority of the variability noted in VO2 in children with CP at a submaximal walking speed.