BACKGROUND: Oral sildenafil and intravenous epoprostenol have independently been shown to be effective in patients with pulmonary arterial hypertension. OBJECTIVE: To investigate the effect of adding oral sildenafil to long-term intravenous epoprostenol in patients with pulmonary arterial hypertension. DESIGN: A 16-week, double-blind, placebo-controlled, parallel-group study. SETTING: Multinational study at 41 centers in 11 countries from 3 July 2003 to 27 January 2006. PATIENTS: 267 patients with pulmonary arterial hypertension (idiopathic, associated anorexigen use or connective tissue disease, or corrected congenital heart disease) who were receiving long-term intravenous epoprostenol therapy. INTERVENTION: Patients were randomly assigned to receive placebo or sildenafil, 20 mg three times daily, titrated to 40 mg and 80 mg three times daily, as tolerated, at 4-week intervals. Of 265 patients who received treatment, 256 (97%) patients (123 in the placebo group and 133 in the sildenafil group) completed the study. MEASUREMENTS: Change from baseline in exercise capacity measured by 6-minute walk distance (primary end point) and hemodynamic measurements, time to clinical worsening, and Borg dyspnea score (secondary end points). RESULTS: A placebo-adjusted increase of 28.8 meters (95% CI, 13.9 to 43.8 meters) in the 6-minute walk distance occurred in patients in the sildenafil group; these improvements were most prominent among patients with baseline distances of 325 meters or more. Relative to epoprostenol monotherapy, addition of sildenafil resulted in a greater change in mean pulmonary arterial pressure by -3.8 mm Hg (CI, -5.6 to -2.1 mm Hg); cardiac output by 0.9 L/min (CI, 0.5 to 1.2 L/min); and longer time to clinical worsening, with a smaller proportion of patients experiencing a worsening event in the sildenafil group (0.062) than in the placebo group (0.195) by week 16 (P = 0.002). Health-related quality of life also improved in patients who received combined therapy compared with those who received epoprostenol monotherapy. There was no effect on the Borg dyspnea score. Of the side effects generally associated with sildenafil treatment, the most commonly reported in the placebo and sildenafil groups, respectively, were headache (34% and 57%; difference, 23 percentage points [CI, 12 to 35 percentage points]), dyspepsia (2% and 16%; difference, 13 percentage points [CI, 7 to 20 percentage points]), pain in extremity (18% and 25%; difference, 8 percentage points [CI, -2 to 18 percentage points]), and nausea (18% and 25%; difference, 8 percentage points [CI, -2 to 18 percentage points]). LIMITATIONS: The study excluded patients with pulmonary arterial hypertension associated with other causes. There was an imbalance in missing data between groups, with 8 placebo recipients having no postbaseline walk assessment compared with 1 sildenafil recipient. These patients were excluded from the analysis. CONCLUSION: In some patients with pulmonary arterial hypertension, the addition of sildenafil to long-term intravenous epoprostenol therapy improves exercise capacity, hemodynamic measurements, time to clinical worsening, and quality of life, but not Borg dyspnea score. Increased rates of headache and dyspepsia occurred with the addition of sildenafil.