The clinical, biochemical, and radiological features of pituitary ACTH-dependent Cushing's syndrome (CS) [Cushing's disease (CD)] are often indistinguishable from those of occult ectopic ACTH-dependent CS (oEAS). We have evaluated, retrospectively, the results of simultaneous bilateral inferior petrosal sinus (IPS) ACTH sampling before and after CRH stimulation in 128 patients with ACTH-dependent CS: 107 patients with CD, 6 with oEAS, 1 with an adrenal adenoma, 1 with a pituitary gangliocytoma, and 1 with Nelson's syndrome; while, in the remaining 12, the source remains unclear. One hundred seven patients received human-sequence CRH (hCRH), and 11 received ovine CRH; another 6 patients underwent stimulation with desmopressin and hCRH, and 4 with desmopressin alone. A successful bilateral IPS catheterization and sampling (IPSCS) rate of 87.5% was obtained only after considerable experience had been gained. Sixty-nine patients with CD underwent successful bilateral IPSCS: the IPS-to-peripheral ratio of plasma ACTH value (IPS/P) rose from 9.5 +/- 1.4 to a maximum ratio of 55.8 +/- 7.5 in 67 patients, after CRH stimulation. The maximum ratio was obtained at 5 min in 60 of the 69 patients with CD; however, all 69 patients obtained a ratio of more than 2, at that time. In contrast, the 6 patients with occult ACTH-secreting neoplasms had a maximal IPS/P ratio of 1.3 +/- 0.16, and this did not change after CRH stimulation. A bilateral IPS/P ratio more than 2, obtained 5 min after CRH stimulation, had a sensitivity of 97% and a specificity of 100% in diagnosing CD. Two patients with proven active CD had an IPS/P ratio of less than 2. After successful bilateral IPSCS, the gradients between the IPS ACTH concentrations [IPS ACTH gradient (IPSG)] of more than 1.4, at 5 min after CRH stimulation, had a sensitivity of 83% in correctly lateralizing the pituitary microadenoma, compared with 72% sensitivity for magnetic resonance imaging (MRI) scanning. Furthermore, when IPSG and MRI findings were contradictory, IPSG was more often correct than MRI scanning. Although oEAS is a relatively uncommon cause of ACTH-dependent hypercortisolism (5.5% in our series), the accurate diagnosis of ACTH-dependent CS and localization of an intrapituitary microadenoma requires bilateral IPSCS with CRH stimulation, provided that the appropriate technical experience is available. hCRH is as effective a secretagogue as ovine CRH, and either may be used. The value of the combination of CRH and desmopressin stimulation requires more detailed investigation.