A theoretical calculation of the total nonaqueous exchangeable hydrogen in protein, carbohydrates, and fat in man has been made. It shows that of the total exchangeable hydrogen in the body 5.22% is located in biochemical components, soluble in body water, containing hydrogen that is exchangeable with the isotope. This value represents a maximum upward distortion of total body water measurements by isotope dilution, due to the maximum possible exchangeability in these molecular conformations. From comparative measurements reported in the literature it is clear that this maximum is not achieved during the short period of time during which tritium-dilution studies are performed. It is the authors' belief that the hard-to-exchange amide hydrogens described by Blout in the protein conformations account for this failure of the isotope to achieve complete exchange in the short time allowed.