A sample of 23 drug-free, normal adult subjects, aged 23 to 50 years, received 1 mg dexamethasone p.o. at midnight. Serum cortisols were obtained at 0800h, 1600h, and 2330h pre- and postdexamethasone. Only 1 of these 23 subjects (4.3%) evidenced nonsuppression, as defined by any postdexamethasone serum cortisol value of greater than 4.0 micrograms/dl. A dose of 0.75 mg dexamethasone was administered to 23 drug-free, normal adult subjects, 20 of whom participated in the above 1 mg trial. Six of these 23 (26.1%) showed nonsuppression at a threshold of 4.0 micrograms/dl. Another 11 normal adults who were taking various prescription medications (e.g., sympathomimetics, nasal decongestants, birth control pills, thyroid hormones) or who were suffering from untreated upper respiratory infections, venereal infections, or allergies were tested with 1 mg of dexamethasone. In this sample, 7 of 13 (53.8%) showed nonsuppression. These findings suggest that: (1) 1 mg of dexamethasone is the lowest effective dose that can be used in diagnostic testing for melancholic depression; (2) a false-positive response to the dexamethasone suppression test (DST) may occur with infections, allergies, or possibly with certain prescription medications. Further studies of the effects of illness and/or medications on DST responses are needed.