Successful and cost-effective prophylaxis against venous thromboembolism (VTE) depends on the availability of safe and effective antithrombotic methods, and the ability to match these to patients according to their level of thromboembolic risk. The last 20 years have seen significant developments in methods of thromboprophylaxis but, despite use of the best available methods, VTE still occurs in a significant proportion of patients. Efforts are now focused both on developing more effective pharmacological and physical methods, and improving assessment of thromboembolic risk in clinical practice, with the goal of avoiding unnecessary prophylaxis in low-risk patients and providing enhanced protection to high-risk patients. Factors known to exert a direct effect on thromboembolic risk include previous VTE, advancing age, surgery, malignancy, immobility, and the presence of thrombophilic states. Other less well-defined risk factors include obesity, pregnancy, and leg varicosities. Various risk assessment models (RAMs) have been devised based on these factors, but most of the resulting risk assessment formulae, some of them based on laboratory test results, have been too complicated to gain acceptance in routine clinical practice. Reflecting the need to refine the use of thromboprophylaxis in clinical practice, a number of RAMs have been developed recently, based on epidemiological evidence. Some models include specific recommendations for prophylaxis, based on evidence from randomized, controlled trials.