There is an understanding that walking with flexed knees contributes to the increased energy cost of walking found in children with neurological conditions. To determine the influence of flexed-knee gait on energy cost of walking in a group of children without neurological abnormality, the gait patterns of 10 normal children were studied using a Vicon system and standard marker set. A telemetric system (Cosmed K2) was then used to measure the oxygen cost of walking of the same children. The tests were repeated restricting the subjects' knee extension bilaterally, using hinged braces, set to 0, 15, 30, and 45 degrees of flexion. Although the braces themselves caused a significant increase in O2 cost (mL/kg/m) (P < 0.05), due to a decrease in walking speed, no further significant increase in oxygen cost was demonstrated regardless of the degree of knee flexion imposed, despite a significant increase in measured hip flexion and ankle dorsiflexion at the 45 degrees of knee flexion setting (P < 0.05). We propose that moderate flexed-knee gait does not of itself cause an increase in the energy cost of walking and that other factors present in the physically disabled child are likely to be implicated.