Pulmonary complications remain the most important cause of postoperative morbidity and mortality. The many advances of modern surgical care over the last 30 years have not appreciably altered the incidence of these complications. Many risk factors have been shown to contribute to this problem, but no studies have examined the impact of preoperative protein depletion on respiratory function and related this to the development of postoperative pulmonary complications. 80 patients (42 men, 38 women, median age of 64 years, with a range of 15-91 years) awaiting major elective gastrointestinal (G.I.) surgery were divided into two categories on the basis of a direct measurement of protein depletion: nonprotein-depleted patients (n = 41, mean protein loss, 2% +/- 1.7 SEM) and protein-depleted patients (n = 39, mean protein loss, 36% +/- 3.5 SEM). There was no significant difference between these two categories in regard to age, height, sex, surgical diagnosis, the presence of chronic lung disease, smoking, proportion of upper abdominal incisions, degree of obesity, the duration of anesthesia, and the use of prophylactic antibiotics and physiotherapy. There was a significant difference between these two categories of patients in regard to respiratory muscle strength (p less than .025), vital capacity (p less than .05), and peak expiratory flow rate (p less than .005). Pneumonia developed in a significantly higher proportion of protein-depleted patients with atelectasis (p less than .05), and their stay in the hospital after surgery was longer (p less than .05). These data show that protein depletion is associated with an impairment of respiratory function, and is in itself a significant risk factor in the development of postoperative pneumonia.