Salivary cortisol measurement is used as a practical surrogate for serum free cortisol. However, parotid tissue harbors 11β-hydroxysteroid dehydrogenase (11β-HSD2) activity converting cortisol to cortisone. This study was designed to assess the impact of parotid 11β-HSD2 activity on the measurement of salivary cortisol. PATIENTS, DESIGN, AND OUTCOME MEASURES: Study participants with changes in circulating corticosteroid-binding globulin (CBG) (±oral contraceptive, functionally CBG null) and controls were studied during adrenal stimulation by ACTH and postoral and iv hydrocortisone administration. Simultaneous serum and saliva samples were collected for the measurement of total serum cortisol (SerF) by immunoassay, and unbound cortisol and cortisone in serum (FreeF and FreeE) and saliva (SalF and SalE) by liquid chromatography-tandem mass spectrometry. ACTH stimulation increased SerF, FreeF, SalF, SalE, but not FreeE in all individuals. SerF significantly decreased after stopping oral contraceptive administration, but FreeF, SalF and SalE remained unchanged. In the hydrocortisone administration study, individual FreeF and SalE curves were nearly identical and SalE closely reflected FreeF in all participants, irrespective of CBG changes. The highest correlation in all (n = 537) matched serum-saliva samples was between SalE and FreeF (r = 0.95, P < 0.0001), and there was no evidence of 11β-HSD2 saturation. Salivary cortisol is a useful surrogate for circulating free cortisol, but its concentration is determined both by serum free cortisol and parotid metabolism to cortisone. We have shown that salivary cortisone closely reflects free serum cortisol after adrenal stimulation and hydrocortisone administration and is unaffected by CBG changes. Salivary cortisone has potential as a useful surrogate for serum free cortisol in research and clinical assessment, and further research in states of chronic glucocorticoid excess is now needed.