Interest in the fundamental mechanisms underlying headache, particularly the pathophysiology of migraine and cluster headache, has lead to the study of the physiology and pharmacology of the trigeminovascular system and its central ramifications. Cats were anaesthetized (60 mg/kg alpha-chloralose, i.p., along with halothane for all surgical procedures) and prepared for physiological monitoring. The animals were placed in a stereotaxic frame and ventilated. A midline craniotomy and C2 laminectomy were performed for access to the superior sagittal sinus and C2 dorsal horn, respectively. The sinus was isolated from the underlying cortex and stimulated electrically after the animals had been paralysed with gallamine (6 mg/kg, i.v.). Units linked to stimulation were recorded with a tungsten-in-glass microelectrode placed in the most caudal part of the trigeminal nucleus, the trigeminocervical complex. Signals from the neurons were amplified, filtered and passed to a microcomputer, where post-stimulus histograms were constructed on-line to analyse the responses to stimulation. Units responded to sagittal sinus stimulation with a typical latency of 8-10 ms. All units studied had a probability of firing of 0.6 or greater. Intravenous injection of the non-competitive N-methyl-D-aspartate receptor antagonist, dizocilpine maleate (4 mg/kg, i.v.), resulted in a substantial and prolonged blockade of firing of units in the trigeminocervical complex. Similarly, administration of the non-N-methyl-D-aspartate excitatory amino acid receptor blocker, GYKI 52466, lead to a dose-dependent inhibition of trigeminovascular-evoked responses in the trigeminocervical complex. These data demonstrate the participation of both N-methyl-D-aspartate- and non-N-methyl-D-aspartate-mediated mechanisms in transmission within the trigeminocervical complex, and suggest a clear preclinical role of glutamatergic mechanisms in primary headache syndromes, such as migraine and cluster headache.