The aim of this study was to perform a national survey on occasionally discovered adrenal masses [adrenal incidentalomas (AI)] under the auspices of the Italian Society of Endocrinology. This multicentric and retrospective evaluation of patients with AI includes 1096 cases collected in 26 centers between 1980 and 1995. Relevant information was obtained by means of a specifically tailored questionnaire. Of the 1096 forms received, 1004 were retained for final analysis. Patients were 420 males and 584 females, aged between 15-86 yr (median, 58 yr). Mass size (computed tomography measurement) ranged from 0.5-25 cm (median, 3.0 cm). Hormonal work-up demonstrated that 85% of the masses were nonhypersecretory, 9.2% were defined as subclinical Cushing's syndrome, 4.2% were pheochromocytomas, and 1.6% were aldosteronomas. Adrenalectomy was performed in 380 patients with removal of 198 cortical adenomas (52%), 47 cortical carcinomas (12%), 42 pheochromocytomas (11%), and other less frequent tumor types. Patients with carcinoma were significantly younger than patients with adenoma (median, 46; range, 17-84; vs. 57, 16-83 yr; P = 0.05). Adenomas were significantly smaller than carcinomas (3.5, 1-15 vs. 7.5, 2.6-25 cm; P < 0.001), and a cut-off at 4.0 cm had the highest sensitivity (93%) in differentiating between benign and malignant tumors. Hormonal work-up of patients with subclinical Cushing's syndrome showed low baseline ACTH in 79%, cortisol unsuppressibility after 1 mg dexamethasone in 73%, above normal urinary free cortisol in 75%, disturbed cortisol rhythm in 43%, and blunted ACTH response to CRH in 55%. Only 43% of patients with pheochromocytoma were hypertensive, and 86% showed elevated urinary catecholamines. All patients with aldosteronoma were hypertensive and had suppressed upright PRA. These results indicate that mass size is the most reliable variable in separating benign from malignant AI. Adrenalectomy should be recommended for AI greater than 4.0 cm because of the increased risk of malignancy, especially in young patients. Endocrine evaluation should be performed in all patients to identify silent states of hormone excess.