To evaluate whether cyclooxygenase constrictor substances can impair nitric oxide-mediated vasodilation in essential hypertension, in seven normotensive subjects (43.3+/-4.1 years; BP, 117+/-6/81+/-2 mm Hg) and seven essential hypertensive patients (47.1+/-5.2 years; BP, 151+/-8/98+/-4 mm Hg) we studied forearm blood flow (strain-gauge plethysmography) modifications induced by intrabrachial acetylcholine (0.15, 0.45, 1.5, 4.5, 15 mu g . 100 ml(-1) . min(-1)) in basal conditions, during infusion of N-G-monomethyl-L-arginine (L-NMMA; 100 mu g . 100 mL(-1) . min(-1)), a nitric oxide synthase inhibitor, or indomethacin (50 mu g . 100 mL(-1) . min(-1)), a cyclooxygenase inhibitor, or simultaneous indomethacin and L-NMMA. In normotensives, vasodilation to acetylcholine was blunted by L-NMMA (maximum flow increase: 671+/-64% and 386+/-42%, respectively; P<.01),and this effect was unchanged by indomethacin. In contrast, in hypertensive patients, vasodilation to acetylcholine (maximum flow increase: 458+/-33%) was unchanged by L-NMMA. Indomethacin significantly (P<.01) increased the response to acetylcholine (maximum flow increase: 635+/-53%) and restored the inhibitory effect of L-NMMA (maximum flow increase: 445+/-36%; P<.01 versus indomethacin alone). In an adjunctive seven normotensives (51.4+/-4.2 years; BP, 114+/-5/79+/-3 mm Hg) and seven essential hypertensives (53.2+/-7.6 years; BP, 153+/-9/100+/-3 mm Hg) we repeated the same protocol by replacing L-NMMA with L-arginine (200 mu g . 100 mL(-1) . min(-1)), the substrate for NO synthase. In normotensives, vasodilation to acetylcholine was increased by L-arginine (maximum flow increase: 539+/-48% and 806+/-61%, respectively) and this effect was unchanged by indomethacin. In hypertensive patients, vasodilation to acetylcholine (maximum flow increase: 339+/-32%) was unchanged by L-arginine but was significantly (P<.01) increased by indomethacin (maximum flow increase: 592+/-38%). Moreover, indomethacin restored the facilitatory effect of L-arginine (maximum flow increase: 804+56%; P<.01 versus indomethacin alone). Therefore, cyclooxygenase inhibition restores nitric oxide-mediated vasodilation in essential hypertension, suggesting that cyclooxygenase-dependent substances can impair nitric oxide production.