The present study set out to investigate the external carotid vascular effects of isometheptene in vagosympathectomised dogs, anaesthetised with pentobarbital. One-minute intracarotid (intra-arterial; i.a.) infusions of isometheptene (10, 30, 100 and 300 microg/min) produced dose-dependent decreases in external carotid blood flow, without affecting blood pressure or heart rate. The vasoconstrictor responses to 100 microg/min and 300 microg/min of isometheptene were clearly attenuated in animals pretreated with reserpine (5,000 microg/kg). Moreover, after prazosin (an alpha1-adrenoceptor antagonist; 100 microg/kg), the responses to isometheptene remained unaltered in untreated as well as reserpine-pretreated dogs. In contrast, the responses to isometheptene were attenuated by rauwolscine (an alpha2-adrenoceptor antagonist; 300 microg/kg) in untreated animals, and were practically abolished in reserpine-pretreated dogs. Further investigation into the specific alpha2-adrenoceptor subtypes, using selective antagonists, showed that BRL44408 (alpha2A) and MK912 (alpha2C) markedly attenuated this response, while imiloxan (alpha2B) was ineffective. The involvement of 5-HT1B and 5-HT1D receptors seems highly unlikely since antagonists at 5-HT1B (SB224289) and 5-HT1D (BRL15572) receptors (both at 300 microg/kg) were ineffective. On this basis, it is concluded that isometheptene-induced canine external carotid vasoconstriction is mediated by both indirect (a tyramine-like action) and direct (acting at receptors) mechanisms, which mainly involve alphaA- and alpha2C-adrenoceptors, while the involvement of alpha1-adrenoceptors seems rather limited.