In 37 consecutive depressed inpatients afternoon plasma cortisol (1500-1520 h) was measured in 3 ways: 1) spontaneously; 2) 2 h after oral administration of 60 mg oxazepam; and 3) 16 h after oral administration of 2 mg dexamethasone. Both oxazepam and dexamethasone caused a significant suppression of cortisol secretion. Spontaneous and suppressed cortisol levels correlated significantly to stress/distress items on the Hamilton Rating Scale for Depression (sum of items 8, 9, 10 and 12), whereas no correlations to age, or type of depression were found. In an earlier similar study of 35 patients both spontaneous and suppressed cortisol levels correlated significantly with age, type and severity of depression as well as with the stress/distress items. Those patients were older, more depressed (HRSD-17) and had higher stress/distress scores compared with the present sample of 37 patients. The consistent finding of a correlation with the stress/distress items suggest that this factor is important in relation to the hypersecretion of cortisol during depression and this may explain the limited diagnostic power of spontaneous and suppressed cortisol levels.