Vasodilator prostaglandins are released in vitro from endothelium during adrenergic stimulation. We hypothesized that indomethacin would block this production in vivo and increase venoconstriction to alpha1- and alpha2-stimulation but not to the nonadrenergic agonist PGF2alpha. Hand vein distension was measured in 24 normal subjects (23.0 +/- 0.5 yr) during local infusions of phenylephrine (8-12,000 ng/min), clonidine (3-7,000 ng/min), or PGF2alpha (1-2,048 ng/min) plus indomethacin (3 microg/min) versus saline on two separate days. Dose-dependent venoconstriction to phenylephrine occurred in all subjects, with a parallel shift to the left with indomethacin (P = 0. 003) and a decrease in the phenylephrine 50% effective dose (1,009 vs. 241 ng/min, geometric means, P = 0.012). Venoconstriction to clonidine was more variable, with most subjects eliciting a biphasic response (initial venoconstriction followed by attenuation). With indomethacin, the dose-response curve was displaced up and to the left (P = 0.005), and peak venoconstriction was increased (51.1 +/- 6.8 vs. 27.2 +/- 5.3% of control, P = 0.018) without a biphasic response. In all subjects, PGF2alpha elicited dose-dependent venoconstriction that was not altered by indomethacin. Thus venous alpha1- and alpha2-stimulation results in release of vasodilator prostaglandins that antagonize the venoconstrictor response. This modulates the sympathetic response of venous smooth muscle and may be important in diseases with endothelial dysfunction.