Recent advances in the understanding of the pathogenesis of cancer have led to the introduction of a variety of biological agents with novel mechanisms of action into clinical trials and even into clinical practice. In particular, tumour-associated neoangiogenesis has become a major target for this new class of antineoplastic agents. Five anti-angiogenic agents (thalidomide, lenalidomide, bevacizumab, sunitinib, sorafenib) have already obtained US Food and Drug Administration approval for clinical use, and many others have entered clinical trials. Many new biological agents with anti-angiogenic properties appear to be associated with an increased risk for thrombosis and, paradoxically, bleeding. Although the mechanisms underlying the increased thromboembolic risk remain ill defined, the main hypothesis is that perturbation of tumour-associated endothelial cells can switch the endothelium from a naturally anticoagulant surface to a prothrombotic surface, thus mediating the activation of systemic coagulation in cancer patients, who are already more susceptible to thromboembolism due to their underlying disease. The toxicity profile differs between the anti-angiogenic agents. Thalidomide, lenalidomide, semaxibin (SU5416) and prinomastat have produced more venous thromboembolic complications, whereas bevacizumab, sunitinib, sorafenib and ZD6126 have been associated with a higher risk of arterial thromboembolism and, in particular, myocardial ischaemia. The observation of these vascular toxicities suggests the need to establish, in randomized clinical trials, the usefulness of thrombosis prophylaxis when anti-angiogenic agents are used in cancer patients, especially when associated with chemotherapy. In addition, careful reporting of haemostatic complications during treatment with new anti-angiogenic drugs is warranted.