Patients who clinically have primary or 'unexplained' pulmonary hypertension are found at autopsy or lung biopsy to have a variety of pulmonary vascular changes, including medial hypertrophy, thrombosis, intimal fibrosis and plexiform lesions. It is not surprising that the haemodynamic response to vasodilators varies widely. In general, the non-specific vasodilators used to treat pulmonary hypertension cause an acute fall in systemic arterial pressure, with an increase in cardiac output and a reduction in pulmonary vascular resistance. Pulmonary arterial pressure usually does not change much but occasionaly drops dramatically. The risk of death in an acute trial of a vasodilator is less than 0.5% in experienced hands. The use of a short-acting vasodilator (e.g., prostacyclin) may indicate the presence or absence of vasoconstriction, the likelihood of fixed structural obstruction to flow and the risk of administering longer-acting vasodilators, and it may give a clue to prognosis. The risk-benefit ratio in the use of vasodilators in the long-term treatment of primary pulmonary hypertension needs to be evaluated by a controlled trial, conducted in those who respond acutely. The role of high-dose calcium channel blocker treatment and multiple drug therapy will also require further study.