Insulin sensitivity was determined before and after elective surgery in 31 otherwise healthy patients undergoing elective surgery for open cholecystectomy (n = 24) or inguinal hernia repair (n = 7) and compared with concomitant plasma concentrations of stress hormones and cytokines. Insulin sensitivity was determined employing the normoglycaemic, hyperinsulinaemic clamp at a plasma insulin concentration of 380 pmol/I and a blood glucose concentration of 4.5 mmol/I. Five of the patients undergoing cholecystectomy were studied again on days 5, 9 and 20 after surgery. Preoperative insulin sensitivity ranged from 2.2 to 14.3 mg/kg/min. All patients exhibited reduced insulin sensitivity on the first postoperative day and the mean value fell from 4.7 (0.4) to 2.7 (0.5) mg/kg/min. More pronounced reductions were found after cholecystectomy. A significant increase was found in plasma concentrationsof interleukin-6 (IL-6) postoperatively as compared to preoperative values. However, no significant changes were seen in the postoperative plasma concentrations of any of the hormones studied in patients undergoing hernia repair, while minor increments were seen in patients undergoing open cholecystectomy. There was a significant (r = 0.50, P = 0.005) linear relationship between the reduction in relative insulin sensitivity and the concomitant plasma levels of IL-6. However, no such relation could be confirmed between the changes in plasma hormone concentrations (neither absolute nor relative changes) and the simultaneous alteration in relative insulin sensitivity. In addition, after including three patients who had undergone ileo-anal pouch construction surgery, the relationship between postoperativeinsulin sensitivity and IL-6 levels was even stronger (r = 0.62, P = 0.001). These results suggest that the immunomodulating effects of endogenous IL-6 is of importance in the acute response after surgery and are associated with the development of insulin resistance, while simultaneous plasma concentrations of stress hormones seem to be less sensitive markers of the degree of postoperative metabolic disturbance.