Objective: Subclinical hypercortisolism (SH) may play a role in several metabolic disorders, including diabetes. No data are available on the relative prevalence of SH in type 2 diabetes (T2D). In order to compare the prevalence of SH in T2D and matched non-diabetic control individuals, we performed a case-controlled, multicenter, 12-month study, enrolling 294 consecutive T2D inpatients (1.7% dropped out the study) with no evidence of clinical hypercortisolism and 189 consecutive age- and body mass index-matched non-diabetic inpatients (none of whom dropped out). Design and methods: Ascertained SH (ASH) was diagnosed in individuals (i) with plasma cortisol after 1 mg overnight dexamethasone suppression >1.8 μg/dl (50 nmol/l), (ii) with more than one of the following: (a) urinary free cortisol >60.0 μg/24 h (165.6 nmol/24 h), (b) plasma ACTH <10.0 pg/ml (2.2 pmol/l) or (c) plasma cortisol >7.5 μg/dl (207 nmol/l) at 24:00 h or >1.4 μg/dl (38.6 nmol/l) after dexamethasone-CRH (serum cortisol after corticotrophin-releasing hormone stimulus during dexamethasone administration) test, and (iii) in whom the source of glucocorticoid excess was suggested by imaging and by additional biochemical tests (for ACTH-dependent ASH). Results: Prevalence of ASH was higher in diabetic individuals than in controls (9.4 versus 2.1%; adjusted odds ratio, 4.8; 95% confidence interval, 1.6–14.1; P = 0.004). In our population the proportion of T2D which is statistically attributable to ASH was approx. 7%. Among diabetic patients, the presence of severe diabetes (as defined by the coexistence of hypertension, dyslipidaemia and insulin treatment) was significantly associated with SH (adjusted odds ratio, 3.8; 95% confidence interval, 1.4–10.2; P = 0.017). Conclusions: In hospitalized patients, SH is associated with T2D.