The 1 mg dexamethasone suppression test (DST) was carried out in patients with a major depressive illness in order to establish whether the results of this test, given before antidepressant or ECT treatment, could predict eventual therapeutic outcome. No significant difference could be detected in overall therapeutic improvement between those patients with a normal or abnormal DST response respectively, based on the 50 ng/ml cortisol cut-off point. However, using 100 ng/ml as a cut-off point it was found that patients with an abnormal DST response (i.e. a post-DST plasma cortisol concentration ⩾ 100 ng/ml) responded significantly better than those who had a normal DST response. These observations were statistically significant for those patients receiving antidepressants and in the combined treatment groups of those patients receiving either antidepressants or ECT.