The authors administered the dexamethasone suppression test (DST) to outpatients, who were free from psychoactive drugs for at least 10 days before the test, with primary affective disorder (N = 60), generalized anxiety disorder (N = 26), panic disorder (N = 22), and agoraphobia with panic attacks (N = 13). With a cortisol value of 5 micrograms/dl considered nonsuppression, there were no significant differences in dexamethasone nonsuppression rates among the diagnostic groups. Scores on the Hamilton Rating Scale for Depression and a melancholia subscale were significantly higher in the depressed group than in the anxiety disorder group. The findings raise questions concerning the specificity of the DST for primary affective disorder in relationship to anxiety disorders.