Deep vein thrombosis of upper limb is a common complication of CVC in patients with cancer. In these patients the risk factors for CVC-related thrombosis are not completely defined. The purpose of this study was to identify the risk factors for CVC-related thrombosis in patients included in a randomized, double-blind, placebo-controlled study aimed at assessing the efficacy and safety of enoxaparin for the prophylaxis of CVC-related thrombosis. CVC-related thrombosis was screened by mandatory venography after 6 weeks of study treatment. A number of patient baseline characteristics were assessed as potential risk factors for CVC-related deep vein thrombosis. Crude associations between risk factors and clinical outcomes were assessed by chi(2) test or Fisher's exact test. Multiple logistic regression analysis was used to identify independent risk factors. A CVC-related thrombosis was found in 50 out of 310 patients (16.1%). At multiple logistic regression analysis, CVC tip misplaced in the upper half of superior vena cava (OR 4.05, 95%CI 1.64-10.02), left-sided CVC insertion (OR 2.29, 95%CI 1.01-5.51) and chest radiotherapy (OR 7.01, 95%CI 1.42-34.66) were independent risk factors for thrombosis. In addition to these risk factors, the presence of distant metastases (OR 9.36, 95%CI 1.53-57.05) increased the risk of thrombosis in patients who received placebo. An inadequate position of the CVC tip, left-sided CVC insertion and chest radiotherapy are independent risk factors for CVC-related thrombosis in cancer patients. Patients with distant metastases have an increased risk for thrombosis in absence of antithrombotic prophylaxis.