This paper reviews the assumptions involved in calculating rates of carbohydrate and fat oxidation from measurements of O2 consumption, CO2 production, and urinary nitrogen excretion. It is shown that erroneous results are obtained in the presence of metabolic processes such as lipogenesis and gluconeogenesis. The apparent rates calculated under these conditions can, however, be interpreted as net rates of "utilization." Thus the apparent rate of carbohydrate oxidation is the sum of the rates of utilization for oxidation and for lipogenesis minus the rate at which carbohydrate is formed from amino acids. The apparent rate of fat oxidation is the difference between the rates of oxidation and synthesis from carbohydrate, so that the apparently negative rates encountered in patients infused with glucose do quantitatively represent net rates of synthesis. Other processes such as synthesis of ketone bodies or lactate at rates greater than their utilization can also disturb the calculations, but the magnitude of the effect can be estimated from appropriate measurements. Methods of correcting the observed gaseous exchange in these circumstances are given.