Resting energy expenditure was measured in a heterogeneous group of weight-losing cancer patients (n = 28) in comparison with noncancer patients with (n = 26) and without (n = 17) weight loss. Energy expenditure was measured in a ventilated hood after an overnight fast. The results are presented in relationship to the nutritional state of the patients. Cancer patients had elevated energy expenditure compared to noncancer patients when compared on several different bases. The slopes of the regressions between the degree of malnutrition and energy expenditure differed between cancer and noncancer patients. Depleted cancer patients had similar resting energy expenditure as predicted for healthy well-nourished individuals, but significantly higher compared with hospitalized malnourished noncancer patients. However, the differences in resting energy expenditure seemed to be less pronounced when very severely depleted cancer and noncancer patients were compared. The results indicate that a small but statistically significantly elevated energy expenditure occurs in a considerable number of cancer patients. It may, therefore, add to weight loss due to anorexia, but anorexia is considered to be quantitatively the more important factor of the two.