Plasma dexamethasone concentrations following oral dexamethasone administration were examined in 78 patients with major depression prior to and during treatment. The test-retest stability of plasma dexamethasone levels within patients was satisfactory with an overall significant positive correlation between tests for each patient. However, significant variability was noted in individual patients. Change in pre-DST cortisol and plasma dexamethasone levels were the two variables, in that order of importance, contributing to change in DST status. In studies examining the clinical utility of serial dexamethasone suppression tests as a guide to recovery from depression, the effect of variability in plasma dexamethasone concentrations should be taken into account.