Reportedly between 8-38% of patients who receive bilateral adrenalectomy for treatment of Cushing's disease will develop Nelson's syndrome. We investigated which factors may predict the development of the syndrome. Eight of 48 patients, bilaterally adrenalectomized for pituitary-dependent Cushing's syndrome 1-30 yr previously, developed Nelson's syndrome 1.5-13 yr (6.6 +/- 4.3 yr) after adrenalectomy. The mean age at adrenalectomy in the group of patients who developed Nelson's syndrome was significantly lower than that in the group without the syndrome (mean +/- SD, 26.0 +/- 6.0 and 35.6 +/- 11.7 yr, respectively; P < 0.02). In the patients adrenalectomized before the age of 35 yr, 8 of 27 (30%) developed Nelson's syndrome, whereas in the patients older than 35 yr, no one did (P < 0.02). No statistically significant differences between the two groups were found in sex ratio, duration of disease before adrenalectomy, or duration of follow-up thereafter. There were no statistically significant differences between the two groups in mean plasma cortisol and ACTH levels before adrenalectomy, cortisol suppressibility after the administration of 8 and 16 mg dexamethasone, or cortisol responses to CRH, TRH, and LH-releasing hormone before adrenalectomy. We conclude that age at the time of adrenalectomy is an important predictive factor for the development of Nelson's syndrome.