Height measurement is required to standardize measures of physical capacity (e.g., pulmonary function) and to adjust drug dosage in the physically disabled. To estimate height in wheelchair-bound patients, we examined the relation of four different upper extremity measurements to height in 119 normal individuals aged 0.5-56 years. Regression analysis was performed for each of the following: (a) interacromioclavicular distance, (b) upper arm, (c) lower arm (cubit), (d) arm span; with height. Pearson correlation coefficients of 0.937, 0.967, 0.975, and 0.989, respectively, were obtained (accuracy ±5%, 90% confidence interval, for upper and lower arm measurements). We conclude that height can be predicted reliably from arm measurements.