Sixty patients undergoing colorectal surgery for malignancy were randomized to receive the anabolic steroid stanozolol (n = 30) or to a control group (n = 30). Patients were further randomized to receive on the first 4 postoperative days a) a standard dextrose-saline regimen (DS), b) an amino acid regimen (AA), or c) a glucose-amino acid-fat regimen (GAF) via a peripheral vein. Fat and carbohydrate oxidation rates were calculated pre- and postoperatively using indirect calorimetry. Postoperative nitrogen balance (NB) in patients receiving amino acids was significantly improved (p less than 0.02) by the administration of stanozolol. Fat and carbohydrate oxidation rates were not significantly affected by stanozolol. Patients in the stanozolol and control AA groups showed a fall in carbohydrate oxidation (p less than 0.01) and a rise in fat oxidation (p less than 0.05) postoperatively, whereas no significant changes in fat and carbohydrate oxidation occurred in the two DS and two GAF groups. Cumulative NB for the first 4 postoperative days was significantly better (p less than 0.01) in the two AA groups than in the two DS groups, due to an improved NB in the two AA groups on the 1st and 2nd days only. Cumulative NB in the two GAF groups was significantly better (p less than 0.01) than in all the other groups. This study shows that stanozolol improves postoperative NB in patients receiving amino acids alone, whereas the provision of a more complete nutritional regimen containing glucose, amino acids, and fat results in a positive NB unaffected by stanozolol.