1. The mechanism of uridine 5'-triphosphate-(UTP-)induced vasoconstriction was studied in the rabbit ear artery. The arteries were incubated and perfused at a constant rate of flow. Vasoconstriction was measured as an increase in perfusion pressure. 2. Noradrenaline, adenosine 5'-triphosphate (ATP) and UTP caused concentration-dependent vasoconstriction. ATP and UTP were approximately equipotent. 3. The vasoconstrictor effect of UTP 300 mumol/l was enhanced by a mixture of atropine, diphenhydramine and methysergide (1 mumol/l each) and not affected by indomethacin 10 mumol/l. 4. Prazosin (0.01-1 mumol/l) and phentolamine (1-10 mumol/l) reduced the vasoconstrictor effect of UTP 300 mumol/l by up to 34%. Prazosin 1 mumol/l failed to diminish the vasoconstrictor effect of UTP 300 mumol/l after the sympathetic nerves had been destroyed with 6-hydroxydopamine. 5. alpha, beta-Methylene-ATP (10-50 mumol/l) elicited transient vasoconstriction. Subsequently, vasoconstrictor responses to ATP 100 or 300 mumol/l were reduced by 88%, whereas responses to UTP 100 mumol/l were enhanced, responses to UTP 300 mumol/l decreased by only 32% and responses to UTP 1000 mumol/l reduced by 74%. After in vitro-denervation with 6-hydroxydopamine or in the presence of phentolamine 1 mumol/l throughout, alpha, beta-methylene-ATP (10-50 mumol/l) reduced the vasoconstrictor effect of UTP 300 mumol/l by 44% and 43%, respectively. 6. We suggest that, in the rabbit ear artery, the non-adrenergic and alpha, beta-methylene-ATP-resistant vasoconstrictor response to UTP is mediated by a separate receptor mechanism, distinct from the P2 purinoceptor.