We tested the hypothesis that dexmedetomidine (Dex) has greater alpha(2)- vs. alpha(1) selectivity than clonidine and causes more alpha(2)-selective vasoconstriction in the human forearm. After local beta-adrenergic blockade with propranolol, forearm blood flow (plethysmography) responses to brachial artery administration of Dex, clonidine, and phenylephrine (alpha(1)-agonist) were determined in healthy young adults before and after alpha(2)-blockade with yohimbine (n = 10) or alpha(1)-blockade with prazosin (n = 9). Yohimbine had no effect on phenylephrine-mediated vasoconstriction but blunted Dex-mediated vasoconstriction (mean +/- SE: -41 +/- 5 vs. -11 +/- 2%; before vs. after yohimbine) more than clonidine-mediated vasoconstriction (-39 +/- 5 vs. -28 +/- 4%; before vs. after yohimbine) (P < 0.02). Prazosin blunted phenylephrine-mediated vasoconstriction (-39 +/- 4 vs. -8 +/- 2%; before vs. after prazosin) but had similar effects on both Dex- (-30 +/- 4 vs. -39 +/- 6%; before vs. after prazosin) and clonidine-mediated vasoconstriction (-29 +/- 3 vs. -41 +/- 7%; before vs. after prazosin) (P > 0.7). Both Dex and clonidine reduced deep forearm venous norepinephrine concentrations to a similar extent (-59 +/- 12 vs. -55 +/- 10 pg/ml; Dex vs. clonidine, P > 0.6); this effect was abolished by yohimbine and blunted by prazosin. These results suggest that Dex causes more alpha(2)-selective vasoconstriction in the forearm than clonidine. The similar vasoconstrictor responses to both drugs after prazosin might be explained by the presynaptic effects on norepinephrine release.