The aim of this study was to estimate the role of Caprini model in forecasting of postoperative venous thromboembolism (VTE) in patients with high surgical risk and to determine patients with extremely high risk of venous thrombosis by using of this scale. It was performed prospective multicenter observational study involving 140 high risk patients who underwent large abdominal (48%) or neurosurgical (52%) intervention. All patients were assessed with the Caprini model and had a mean score of 9.5±2.7 (range: 5-15). Complex postoperative prevention of venous thromboembolic complications included elastic compression of low extremities and standard prophylactic doses of direct anticoagulants. Duplex ultrasonic scanning was performed during the first 12 hours after surgery and then every 3-5 days until discharge to assess the lower limbs venous system. If venous thrombosis was observed pulmonary embolism has been excluded by using of echocardiography, planar perfusion scintigraphy of the lungs, SPECT-imaging with computed tomography. Venous thrombosis was diagnosed in 39 (28%) patients. The incidence of VTE was 1.9% in the lowest tertile of the Caprini score (5-8); it was 26.1% in the middle tertile (score 9-11) and 65% in the upper tertile (score 12-15) (p<0.001). The area under ROC curve was 0.874 (95% CI 0.81 to 0.94) and Caprini score 11 was a cut-off point that provided the highest sensitivity and specificity. 2 (2.6%) of 77 patients with score less than 11 had venous thrombosis. In contrast, 37 (58.7%) of 63 patients with Caprini score 11 and more had venous thrombosis (p<0.001). There is significant correlation between Caprini scores and frequency of postoperative venous thromboembolism in patients with high surgical risk. Score 11 and more can identify patients with extremely high risk. These patients need for more effective prevention.