Objective. Several smaller studies on adrenal incidentalomas (AI) suggested an association between autonomous cortisol secretion (ACS) and mortality (Di Dalmazi Lancet Diabetes Endocrinol 2014, Debono J Clin Endocrinol Metab 2014, Patrova Endocrine 2017). However, a recent meta-analysis (9 studies, 1356 patients) could not confirm these findings (Elhassan Ann Intern Med 2019). Aim. To investigate the effects of ACS on mortality, prevalence of cardiovascular (CV) risk factors, and (CV) morbidity, in a representative cohort of AI. Design. Retrospective observational study conducted at 27 ENS@T centers from 15 countries. Methods. Inclusion criteria: AI diagnosed 1996-2015, 1 mg dexamethasone suppression test, follow-up (FU) of ≥36 months, known survival status. Exclusion criteria: clinically relevant adrenal hormone excess (i.e. Cushing’s syndrome, pheochromocytoma, primary hyperaldosteronism), known malignancy. Patient stratification: serum cortisol after dexamethasone (>5 µg/dl, ACS; 1.9-5 µg/dl, possible ACS (PACS); ≤1.8 µg/dl, non-functioning adenoma (NFA)). Definition of CV events (CVE): hospitalization due to myocardial infarction and related interventions (PTCA, surgical bypass), stroke, deep vein thrombosis, pulmonary embolism. Results. 3640 patients (57% NFA, 36% PACS, 7% ACS) were considered eligible: 64% females; median age 61 years (range 18-91); median FU 84 months (36-277) (distribution between subgroups n.s.). 352 patients died during FU. Age- and sex adjusted overall survival was significantly reduced in patients with PACS (HR 1.55; 95%CI 1.24-1.94) and ACS (1.84; 1.29-2.61). Prevalence of CV risk factors were significantly higher in PACS and ACS than in NFA (hypertension: 72, 73, 57%, p<0.0001; dyslipidemia: 42, 49, 35%, p<0.0001; diabetes: 22, 25, 17%, p<0.0001) When adjusted to relevant confounders (i.e. age, sex, CV risk factors), time to first CVE was shorter in PACS (HR 1.36; 1.07-1.73) and ACS (HR 1.62; 1.10-2.40) compared to NFA. Conclusion. PACS and ACS are associated with increased overall mortality and CV morbidity. However, to prove causality a large randomized intervention trial is required.