Background: Anthropometric measurements are widely used to reliably quantify body composition and to estimate risks of overweight in healthy subjects and in patients. However, information about the reliability of anthropometric measurements in subjects with severe intellectual and sensory disabilities is lacking. Objective: The purpose of this study was to determine the feasibility and the test-retest reliability of body composition measures in subjects with severe intellectual and sensory disabilities. Method: The study population consisted of 45 subjects with severe intellectual and sensory disabilities. Body mass index, waist circumference, skin folds and tibia length were measured. Reliability was assessed by Wilcoxon signed rank test, limits of agreement (LOA) and intraclass correlation coefficients. The outcomes were compared with values provided by the World Health Organization. Results: There were no significant differences between test and retest (P less than 0.05). For the skinfold measurements, however, the LOA was insufficient. Intraclass correlation coefficients for all variables, except skinfold measurements, were 0.90 or above. Conclusion: Test-retest reliability and feasibility for all measurements are acceptable in subjects with severe intellectual and sensory disabilities. Skinfold measurements, however, could not be reliably performed in these subjects. Measuring tibia length and using the determined formula to calculate body height from tibia length is a reliable alternative for measuring body height. Although measuring the body height of subjects with severe disabilities was feasible, measuring tibia length was more feasible.