Systemic infection induces an increase in plasma cortisol which accords approximately with illness severity. However, both basal and synthetic ACTH stimulated cortisol levels are not strong predictors of mortality. Moreover, plasma cortisol levels do not readily define those patients who have been clinically observed to respond, with respect to blood pressure elevation, to exogenous hydrocortisone. It is likely that free cortisol, accounting for 6-20% of circulating total (bound plus free) cortisol has most of the life-saving effects on circulation and metabolism in severe sepsis, as corticosteroid-binding globulin bound and albumin-bound cortisol have reduced access to tissues. In addition, sepsis reduces CBG and albumin levels, hence blunting the effect of increasing illness severity on total cortisol. Our recent studies suggest that free cortisol correlates more closely to sepsis severity than total cortisol and that free cortisol levels can be estimated using the plasma CBG and total cortisol, obviating the need for direct free cortisol measurement. Studies directed at determining if free cortisol is a better guide than total cortisol to the need for hydrocortisone supplementation may be of value.