The possibility that adrenoceptors of the .alpha.-subtype mediate vasoconstriction of arteries in response to administered catecholamines was investigated. Clonidine, which in vitro, stimulates .alpha.2-adrenoceptors, was infused into a brachial artery in 12 subjects (0.48 .mu.g/min per 100 ml tissue for 3 min). Afterward, prazosin was infused intraarterially in the 1st 6 subjects (0.5 .mu.g/min per 100 ml for 10 min) and in the remaining subjects, phentolamine was infused (0.12 .mu.g/min per 100 ml) for 10 min. Subsequently the clonidine infusion was repeated. Clonidine decreased forearm blood flow from 3.5 .+-. 0.52 to 1.8 .+-. 0.32 in the 1st 6 subjects and from 4.2 .+-. 0.84 to 2.7 .+-. 0.61 ml/min per 100 ml in the other subjects. .alpha.1-Adrenoceptor blockade by prazosin increased forearm blood flow by 122.7 .+-. 33.8% and combined .alpha.1- and .alpha.2-blockade by phentolamine by 127.2 .+-. 29.9%, indicating much the same degree of postjunctional .alpha.-adrenoceptor blockade. .alpha.2-Adrenoceptor-mediated vasoconstriction by clonidine was abolished after phentolamine (9.1 .+-. 2.29 and 9 .+-. 2.51 ml/min per 100 ml), but was still present after prazosin (7.8 .+-. 1.7 and 4.8 .+-. 1.6 ml/min per 100 ml). Apart from the classical .alpha.1-adrenoceptor, there apparently is a 2nd type of adrenergic receptor on smooth muscle cells that can mediate vasoconstriction, resembling the .alpha.2-adrenoceptor pharmacologically.