Objective: To assess the structural and functional characteristics of pulmonary arteries by intravascular ultrasound (IVUS) in the setting of primary pulmonary hypertension, and to correlate the ultrasound findings with haemodynamic variables and mortality at follow up. Design: Prospective observational study. Setting: University hospital (tertiary referral centre). Patients: 20 consecutive patients with primary pulmonary hypertension (16 female; mean (SD) age, 39 (14) years). Methods: Cardiac catheterisation and simultaneous IVUS of pulmonary artery branches at baseline and after infusion of epoprostenol. Results: 33 pulmonary arteries with a mean diameter of 3.91 (0.80) mm were imaged, and wall thickening was observed in all cases, 64% being eccentric. Mean wall thickness was 0.37 (0.13) mm, percentage wall area 31.0 (9.3)%, pulsatility 14.6 (4.8)%, and pulmonary/elastic strain index 449 (174) mm Hg. No correlation was observed between IVUS findings and haemodynamic variables. Epoprostenol infusion increased pulsatility by 53% and decreased the pulmonary/elastic strain index by 41% (p = 0.0001), irrespective of haemodynamic changes. At 18 (12) months follow up, nine patients had died. A reduced pulsatility and an increased pulmonary/elastic strain index were associated with increased mortality at follow up ( Conclusions: IVUS demonstrated pulmonary artery wall abnormalities in all patients with primary pulmonary hypertension, mostly eccentric. The severity of the changes did not correlate with haemodynamic variables, and epoprostenol improved pulmonary vessel stiffness. There was an association between impaired pulmonary artery functional state as determined by IVUS and mortality at follow up. P rimary pulmonary hypertension is a life threatening disease characterised by a progressive increase in pulmonary blood pressure that often leads to right ventricular failure and death.1 Median survival is 2.8 years from the time of diagnosis, and mortality reaches 65% at three years of follow up.2 Calcium channel blockers, warfarin, and prostacyclin have improved the prognosis, but the three year mortality has remained as high as 50%. The diagnosis of primary pulmonary hypertension is based on clinical and haemodynamic data, and prognosis is determined by the alterations in haemodynamic variables (mean pulmonary artery pressure, cardiac output, mean right atrial pressure).The assessment of pulmonary artery morphology in primary pulmonary hypertension has been limited to pulmonary angiography and to the histological study of lung samples obtained at biopsy. Pulmonary angiography, which is not free of complications in these cases, only shows the vessel lumen and provides no information about vessel wall abnormalities. Histological evaluation of lung biopsies provides a valuable quantitative and qualitative description of the pulmonary wall changes, but remains a static in vitro examination without functional assessment and requires a thoracotomy. Intravascular ultrasound (IVUS) has been validated as a reliable method for...