Many children have contractures and/or deformities that preclude positioning in a fully supine position. The purpose of this study was to evaluate the effects of "poor" positioning on the assessment of body composition with dual X-ray absorptiometry (DXA) in thirty-seven normal child volunteers ages 3-16 yr. Multiple whole-body DXA scans of each child were performed: duplicate scans in the correct fully supine position, two scans while simulating different positions typical of children with contractures, and a scan while positioned in the full lateral position as a "worst-case" scenario. Also evaluated were the precision of duplicate measures in the altered positions, the effect of knee flexion contractures, and the impact of metallic orthopedic fixation devices. Errors in body composition assessment did occur from "poor" positioning. In those positions simulating children with contractures, the mean errors were 4-6% for measures of bone mineral content, 1-3% for lean body mass, and 5-11% for fat mass. Measures in the correct fully supine position and the contracted positions were highly correlated. The errors obtained by altering position were small. If errors of this magnitude are of significance, then corrective equations could be utilized to improve accuracy. Precise and reasonably accurate measures of body composition can be obtained with DXA in children with disabilities and deformities that preclude fully supine positioning.