It is a long held belief that weight loss is a basic indicator of surgical risk. Many experienced surgeons, however, think otherwise. We have investigated the proposition that weight loss is a risk factor for postoperative complications but only when associated with clinically obvious physiologic impairment. Before major surgery, 102 patients had a careful history taken to ascertain if there had been recent weight loss and a reduction in the capacity for activity. Physical examination included assessment of mood, skeletal muscle function, respiratory muscle function, and wound healing. Plasma albumin was also measured. Using this information the patients were placed into one of three groups. Group I (N = 43) were normal, group II (N = 17) had weight loss greater than 10% but no clinical evidence of physiologic impairment, and group III (N = 42) had weight loss greater than 10% with clear evidence of dysfunction of two or more organ systems. The patients in group III had significantly more postoperative complications (p less than 0.05). They also had more septic complications (p less than 0.02) including a higher incidence of pneumonia (p less than 0.05) and a longer hospital stay (p less than 0.05) than patients in each of the other two groups. Objective measurements of body stores of protein and liver, and psychologic, respiratory, and skeletal muscle function, confirmed the validity of the clinical classification into the risk groups. The results demonstrate that weight loss is a basic indicator of surgical risk in modern practice providing it is associated with clinically obvious impairment of organ function. They suggest that adequate body protein stores are necessary for normal body function and for minimizing the risks of surgery.