Cortisol bound to corticosteroid binding globulin (CBG) contributes up to 90% of the total cortisol concentration in circulation. Therefore, changes in the binding kinetics of cortisol to CBG can potentially impact on the concentration of free cortisol, the only form that is responsible for the physiological function of the hormone. When CBG is cleaved into elastase-cleaved CBG (eCBG) by the activity of neutrophil elastase, its affinity for cortisol is reduced. Therefore, when eCBG coexists with intact CBG (iCBG) in plasma, the calculation of free cortisol concentration based on the formulae that considers only one CBG pool with the same affinity for cortisol may be inappropriate. In this study, we developed in vivo and in vitro models of cortisol partitioning which considers two CBG pools, iCBG and eCBG, with different affinities for cortisol, and deduce a new formula for calculating plasma free cortisol concentration. The formula provides better estimates of free cortisol concentration than previously used formulae when measurements of the concentrations of the two CBG forms are available. The model can also be used to estimate the affinity of CBG and albumin for cortisol in different clinical groups. We found no significant difference in the estimated affinity of CBG and albumin for cortisol in normal, sepsis and septic shock groups, although free cortisol was higher in sepsis and septic shock groups. The in vivo model also demonstrated that the concentration of interstitial free cortisol is increased locally at a site of inflammation where iCBG is cleaved to form eCBG by the activity of elastase released by neutrophils. This supports the argument that the cleavage of iCBG at sites of inflammation leads to more lower-affinity eCBG and may be a mechanism that permits the local concentration of free cortisol to increase at these sites, while allowing basal free cortisol concentrations at other sites to remain unaffected.