1 We describe the effects of pretreatment with 5-hydroxytryptamine (5-HT) receptor agonists and antagonists on neurogenically-mediated plasma protein extravasation (["25I]-albumin) in rat dura mater and in extracranial tissues (temporalis muscle fascia, conjunctiva, eyelid and lip) induced by electrical stimulation of the right trigeminal ganglion. 2 Leakage of [25I]-bovine serum albumin from blood vessels in dura mater following high intensity stimulation (1.2 mA, Sims, 5Hz for 5min) was significantly reduced by the intravenous administration of drugs active at 5-HT receptors with some selectivity for the 5-HT1 receptor subtypes: 5-carboxamidotryptamine (5-CT) (threshold dose, 1ngkg-1); 5-benzyloxytryptamine (5-BT) (10, 30 or 100augkg-1); 8-hydroxydipropylaminotetralin (8-OH-DPAT) (300pigkg-1); and as previously reported, sumatriptan (lOOpugkg -), dihydroergotamine (DHE) (SOugkg 1), ergotamine tartrate (lOO1ugkg 1) and chronically administered methysergide (1 mgkg-1).3 The putative 5-HT receptor antagonist, metergoline lOOpgkg-1, inhibited partially the effect of sumatriptan in dura mater providing additional evidence for a 5-HT1 receptor subtype-mediated mechanism, although it was not effective against 5-CT (1 ng kg-1). Methiothepin (300,ug kg-1) did not affect the response to sumatriptan. When administered at high concentrations (1 mg kg 1) methiothepin and metergoline decreased plasma protein extravasation in rat dura mater. 4 Pretreatment with the 5-HT2 receptor antagonists pizotifen, 300pugkg 1, or ketanserin, 300,ugkg ', or the 5-HT3 receptor antagonists MDL 72222, 300,ugkg-1, or ICS 205-930, 300pgkg-1, did not affect plasma protein leakage following electrical trigeminal stimulation. Blockade by sumatriptan of plasma protein extravasation was not inhibited by pizotifen (300,ug kg-1) or MDL 72222 (300pg kg-').5 The 5-HT receptor(s) mediating this response were present only on intracranial tissues innervated by the trigeminal nerve; plasma protein extravasation in extracranial tissues was not blocked by pretreatment with the equivalent or higher concentrations of the above drugs following low intensity trigeminal stimulation (0.1 mA, 5 ms, 5 Hz). 6 The putative 5-HT receptor(s) mediating this response were not present on sympathetic fibres innervating dura mater since unilateral removal of the superior cervical ganglion did not prevent the development of plasma protein extravasation nor did it affect the blockade by sumatriptan IOOpug kg-'. 7 The above pharmacological data suggest that intracranial vessels possess 5-HT receptor(s) which are coupled to inhibition of neurogenically-mediated plasma protein extravasation. These receptors cannot be detected on extracranial cephalic blood vessels innervated by the trigeminal nerve, although available evidence strongly suggests that the 5-HT receptors reside on perivascular trigeminal nerve fibres. The rank order of effective doses (threshold concentrations; 5-CT < 5-BT < DHE < sumatriptan < 8-OHDPAT) is most consistent with a 5-HTlB-or 5-HTlD-mediated response, ...