Serum total cortisol has traditionally been used for the interpretation of tests of the hypothalamic-pituitary-adrenal axis. Approximately 80% of total cortisol is bound to cortisol-binding globulin (CBG), and variation in CBG significantly affects serum total cortisol levels. Reliable assessment of hypothalamic-pituitary-adrenal axis reserve is difficult in severely ill patients, because CBG falls substantially during the acute phase response. The free cortisol index (FCI), defined as the ratio of total cortisol/CBG, correlates well with serum free cortisol. We evaluated the FCI in the context of severe stress and the acute phase response by measuring total cortisol and CBG pre- and postoperatively in 31 patients undergoing major elective surgery. Serum total cortisol increased by 55% from 453 +/- 35.2 (mean +/- SEM) nmol/liter (range, 88-882) to 700 +/- 47.2 (range, 294-1631) nmol/liter. Serum CBG decreased by 30% from 45 +/- 1.7 (range, 26.6-64.1) to 31.4 +/- 1.62 (range, 16.1-51.9) mg/liter, but FCI increased by 130% from 10 +/- 0.8 (range, 2-18) to 23 +/- 1.7 (range, 13-58) nmol/mg. In seven patients (23%), postoperative serum total cortisol was less than 500 nmol/liter, but their postoperative CBG levels were significantly lower than levels in the rest of the group (P < 0.01). However, there was no difference in the FCI between this subgroup and the rest of the group. This study demonstrates the importance of CBG measurement and the calculation of FCI for the interpretation of serum total cortisol in situations where CBG changes significantly.