Salivary cortisol was measured as an alternative to serum cortisol as a marker for adrenocortical function following insulin tolerance test, corticotropin-releasing-hormone stimulation and adreno-corticotrophic hormone stimulation. During insulin tolerance test and corticotropin-releasing-hormone stimulation adreno-corticotrophic hormone was also measured. The tests were performed on healthy control subjects as well as on patients under investigation for various disturbances in the hypothalamic-pituitary-adrenocortical axis (insulin tolerance test: 3 controls on two occasions and 14 patients; corticotropin-releasing-hormone stimulation: 4 controls and 18 patients; adreno-corticotrophic hormone stimulation: 6 controls and 10 patients). Five patients underwent both insulin tolerance test and corticotropin-releasing-hormone stimulation. Using criteria for adequate cortisol response in serum, the patients were classified as good or poor responders. In 42 of the 45 tests performed the same conclusion as to cortisol status was drawn when based on serum and salivary cortisol responses. In healthy subjects and good responders the mean cortisol relative increase was greater in saliva than in serum in all three tests (p < 0.05). Characteristic of the results for the insulin tolerance test was a significant initial mean decrease (p < 0.05), not found in serum, and the highest observed salivary cortisol value was delayed for at least 30 minutes compared to that in serum. Plasma adreno-corticotrophic hormone correlated significantly with the cortisol concentrations determined 15 minutes later in serum (r = 0.54-0.64) and in saliva (r = 0.76-0.85). The more pronounced cortisol response in saliva than in serum and its closer correlation with adreno-corticotrophic hormone offer advantages over serum cortisol, suggesting salivary cortisol measurement may be used as an alternative parameter in dynamic endocrine test.