Acute pulmonary embolism with haemodynamic instability has a high mortality rate. Death results from an acute increase in right ventricular afterload, and the commonly held view is that mechanical obstruction of the pulmonary vascular bed is largely responsible for this increase. In accordance, recent treatment guidelines for severe pulmonary embolism focus exclusively on interventions aimed at relieving this mechanical obstruction, either by thrombolysis or (catheter) embolectomy. However, there is evidence to indicate that vasoconstriction is a very important contributor to the initial increase in pulmonary vascular resistance after pulmonary embolism. This is consistent with the observation that the degree of mechanical obstruction correlates at best poorly with haemodynamic manifestations. Thromboxane A(2) and serotonin are probably mainly responsible for pulmonary vasoconstriction. Cyclooxygenase inhibitors and serotonin antagonists have been shown in animal experiments to attenuate the haemodynamic response to acute pulmonary embolism and to reduce mortality. In addition, reports of a favourable response to pulmonary vasodilators in animals and in humans with acute severe pulmonary embolism have been published. In this paper, it is argued that we may need to reconsider our current therapeutic approach to patients with acute severe pulmonary embolism. Antagonising pulmonary vasoconstrictive mediators or administering pulmonary vasodilators may prove to be life-saving interventions in these patients.