Venous thromboembolism (VTE) is frequently associated with hypercortisolemia. This retrospective single-center study aimed to clarify the significance of plasma D-dimer levels for VTE screening in patients with subclinical or overt Cushing's syndrome (soCS). A total of 72 consecutive treatment-naïve patients with soCS diagnosed at Kobe University Hospital between 2002 and 2014 were enrolled. Patients with both lower extremity ultrasound and D-dimer measurement data (n = 19) were recruited in study 1 and divided into 2 groups, a deep vein thrombosis (DVT) (-) group (n = 12) and DVT (+) group (n = 7) for a comparison of the associated factors. The age and D-dimer levels were higher in the DVT (+) group than in the DVT (-) group (p = 0.04 and 0.02, respectively). A receiver operating characteristic analysis found that D-dimer level ≥2.6 μg/mL correlated with the presence of DVT (sensitivity, 100%; specificity, 91.7%). Next, patients with D-dimer measurement data (n = 36) were recruited in study 2 and divided into 2 groups according to D-dimer level: D-dimer (-), <1 μg/mL group (n = 23) and D-dimer (+), ≥1 μg/mL group (n = 13); the groups were compared with respect to various VTE-related risk factors. A logistic regression analysis revealed that elevated cortisol level after low-dose dexamethasone suppression was a significant risk factor for D-dimer elevation (OR = 1.21, p = 0.02). In conclusion, these data demonstrate that a D-dimer level ≥2.6 μg/mL is an indicator of DVT in treatment naïve patients with soCS and suggests that relatively high autonomous cortisol secretion may be associated with thrombus formation.