BACKGROUND Decreased nitric oxide (NO) is considered an important pathogenetic mechanism in pulmonary arterial hypertension (PAH), but clear evidence is lacking. OBJECTIVE We used multiple techniques to assess endogenous NO in 10 patients with untreated PAH (8 idiopathic and 2 anorexigen-associated PAH) and 12 control subjects. METHODS After a nitrite/nitrate-restricted diet, NO metabolites (NOx) were assayed in 24-hour urine collections and exhaled NO (FE(NO)) determined at multiple expiratory flows. Analysis of the relation between FE(NO) and flow allowed derivation of three flow-independent parameters: airway wall concentration (C(W)), diffusing capacity (D(NO)), and alveolar concentration (C(A)). Seven patients underwent follow-up testing after 3 months of bosentan treatment. RESULTS At baseline, FE(NO) was markedly decreased at the two lowest expiratory flows in PAH: 21 +/- 4 versus 36 +/- 4 ppb at 18 ml/second and 11 +/- 2 versus 17 +/- 2 ppb at 50 ml/second, for subjects with PAH and control subjects, respectively (p < 0.05). C(W) was 33 +/- 11 ppb in subjects with PAH versus 104 +/- 34 in control subjects (p = 0.04). Urinary NOx was also reduced in PAH (42 +/- 6 microM NOx/mM creatinine versus 62 +/- 7 in control subjects; p = 0.04). After bosentan, FE(NO), C(W), and urine NOx increased to control values (p < 0.05). Exclusion of the two anorexigen cases did not alter these results. CONCLUSIONS FE(NO) at low expiratory flows was decreased in PAH due to reduced C(W). Bosentan reversed these abnormalities, suggesting that suppression of NO in PAH may have been caused by endothelin.