The study was undertaken to evaluate whether the Physiological Cost Index (PCI) can be used as a reliable index of efficiency of gait and as an outcome measure in cerebral palsy (CP). Physiological Cost Index was calculated in normal subjects by recording the heart rate manually and with electrocardiograph recording, and the values compared. In another group of subjects, PCI was calculated after they walked 3 different distances (50, 100, and 150 m). The PCI of normal children and children with CP was then estimated by manual recording of the pulse, with the children walking 50 m indoors and 50 m on an uneven surface outdoors. The reproducibility of calculation of PCI was evaluated. The PCI value of each patient was compared to the corresponding Functional Mobility Score. In a group of children with CP, PCI was calculated before and after therapeutic intervention. The PCI values were comparable with either method of heart rate measurement and for the 3 distances walked. The reproducibility of measurement of PCI was satisfactory (Intraclass Correlation Coefficients, 0.80-0.88). The PCI of normal children was 0.1 beats per meter, whereas children with CP had 6 times higher values of PCI, with the highest values in children with a crouch gait. In normal children, 10% greater PCI values were noted when they walked outdoors compared to a 100% increase in children with CP. The higher the PCI values, the lower the Functional Mobility Scores. Therapeutic interventions altered PCI values, and interventions that effectively reduced energy consumption could be identified. We conclude that PCI may be used as a reliable outcome measure of gait efficiency in children with CP.