The vasoconstrictor effects of 5-hydroxytryptamine (5-HT) in the internal carotid bed of anaesthetised dogs with bilateral vagosympathectomy are mainly mediated by both 5-HT1B and 5-HT2 receptors. The blockade of this vasoconstrictor effect of 5-HT by the combined use of the antagonists, GR127935 (5-HT1B/1D) and ritanserin (5-HT2), unmasks a dose-dependent vasodilator effect of 5-HT, but not of sumatriptan. Therefore, the present study set out to analyse the pharmacological profile of this vasodilator 5-HT receptor in the internal carotid bed of vagosympathectomized dogs systematically pretreated with intravenous (i.v.) injections of GR127935 (30 microg/kg) and ritanserin (100 microg/kg). One-minute (1-min) intracarotid (i.c.) infusions of 5-HT (0.1-10 microg/min), 5-carboxamidotryptamine (5-CT; 0.01-0.3 microg/min), 5-methoxytryptamine (5-MeO-T; 1-100 microg/min) and acetylcholine (ACh; 0.003-0.1 microg/min) resulted in dose-dependent increases in internal carotid blood flow (without changes in blood pressure or heart rate) with a rank order of agonist potency of ACh > 5-CT >> 5-HT > or =5-MeO-T. The internal carotid vasodilator responses to 5-HT, 5-CT and 5-MeO-T, which remained unaffected after saline (0.03 ml/kg and 0.1 ml/kg, i.v.), were specifically and dose-dependently blocked by i.v. administration of lisuride (10 microg/kg and 30 microg/kg), clozapine (1000 microg/kg), mesulergine (300 microg/kg and 1000 microg/kg) and LY215840 (300 microg/kg and 1000 microg/kg) with the following apparent rank order of potency: lisuride >> mesulergine = LY215840 > or = clozapine. The above results suggest that the 5-HT receptor mediating internal carotid vasodilatation in vagosympathectomized dogs pretreated with GR127935 and ritanserin is operationally similar to other 5-HT7 receptors mediating vascular and non-vascular responses.