To see whether D-Dimer levels can identifying patients at high risk of venous thrombotic events and establish the best benefit/risk-of-bleeding ratio. Current guidelines do not recommend routine prophylaxis against venous thromboembolism (VTE) in cancer patients receiving chemotherapy, but the risk increases about 6.5-fold because of this treatment. D-dimer was measured at baseline in 124 cancer patients scheduled for their first chemotherapy. VTE events, including symptomatic episodes of deep vein thrombosis or pulmonary embolism or both, were recorded during the first 6 months of therapy, and asymptomatic deep vein thrombosis was revealed by compression ultrasonography at baseline and after 90 and 180 days. During follow-up, there were 11 episodes of VTE (8.9%). Mean D-dimer values were higher in patients with VTE (2195 +/- 1382 vs. 695 +/- 1039 ng/ml, (P < 0.001). On grouping D-dimer values in tertiles, only 2.4% (confidence interval, 0.9-5.7%) in the first (<262 ng/ml) and second tertiles (262-650 ng/ml) suffered a deep vein thrombosis/pulmonary embolism event as compared with 22% (confidence interval, 9-34%) in the third (>650 ng/dl) (P = 0.003). The VTE-free interval was significantly shorter in the third tertile than in the first (P = 0.0218, log-rank test; relative risk for third vs. first tertile, 11.0; 95% confidence interval, 1.4-81.3; P = 0.0033). Multivariate analysis found that only baseline D-dimer concentrations were correlated with the subsequent development of VTE. Baseline D-dimer values in cancer patients scheduled for chemotherapy might be used to select those at low risk of VTE, most likely to be safe without prophylaxis.