Cushing's syndrome requires a screening test of high sensitivity, followed by biochemical evaluation of the source of the tumor when the cause is ACTH dependent. The high-dose dexamethasone suppression test is still in common use as an aid in differential diagnosis, although its value has been queried. We have routinely used the low-dose dexamethasone suppression test for many years in the diagnosis of Cushing's syndrome but noticed that patients with pituitary-dependent Cushing's syndrome or Cushing's disease, usually showed some degree of suppression of their serum cortisol, compared to those with the ectopic ACTH syndrome. We therefore analyzed retrospectively the serum cortisol responses during the low-dose dexamethasone suppression test and the high-dose dexamethasone suppression test in 245 patients with ACTH-dependent Cushing's syndrome and compared the diagnostic utility of each test either alone or in combination with a standard test using CRH. Evaluation of the serum cortisol response at 24 and 48 h during the low-dose dexamethasone suppression test correctly identified 98% of patients with ACTH-dependent Cushing's syndrome and distinguished between pituitary and ectopic causes with a sensitivity of 82% and a specificity of 79%. In the same patients, the serum cortisol response to the high-dose dexamethasone suppression test had a slightly higher sensitivity (91%) and specificity (80%). However, the combined criteria of a more than 30% suppression of serum cortisol during the low-dose dexamethasone suppression test and/or a more than 20% increase in the CRH test had a significantly higher sensitivity (97%) and specificity (94%) than either the high-dose dexamethasone or the CRH tests alone in the differential diagnosis of ACTH-dependent Cushing's syndrome. It produced equivalent information to that when high-dose and CRH test results were combined. We therefore conclude that in our patient series, the serum cortisol response during the low-dose dexamethasone suppression test is highly sensitive in diagnosing Cushing's syndrome and, combined with the results of the serum cortisol response to the CRH test, offered a safe and cost-effective test in the differential diagnosis of ACTH-dependent Cushing's syndrome. There does not appear to be any necessity for retaining the high-dose dexamethasone suppression test in this diagnostic work-up.