1. Voltage-dependent K+ channels are present in cerebral arteries and may modulate vascular tone. We used 200 microM 4-aminopyridine (4-AP), thought to be a relatively selective inhibitor of voltage-dependent K+ channels at this concentration, to test whether activation of these channels may influence baseline diameter of the basilar artery and dilator responses to nitric oxide (NO) and cyclic GMP in vivo. 2. Using a cranial window in anaesthetized rats, topical application of 4-AP to the basilar artery (baseline diameter = 240+/-5 microm, mean +/- s.e.mean) produced 10+/-1% constriction. Sodium nitroprusside (a NO donor), acetylcholine (which stimulates endothelial release of NO), 8-bromo cyclic GMP (a cyclic GMP analogue), cromakalim (an activator of ATP-sensitive K+ channels) and papaverine (a non-NO, non-K+ channel-related vasodilator) produced concentration-dependent vasodilator responses that were reproducible. 3. Responses to 10 and 100 nM nitroprusside were inhibited by 4-AP (20+/-4 vs 8+/-2% and 51+/-5 vs 33+/-5%, respectively, n=10; P<0.05). Responses to acetylcholine and 8-bromo cyclic GMP were also partially inhibited by 4-AP. In contrast, 4-AP had no effect on vasodilator responses to cromakalim or papaverine. These findings suggest that NO/cyclic GMP-induced dilator responses of the basilar artery are selectively inhibited by 4-aminopyridine. 4. Responses to nitroprusside were also markedly inhibited by 10 microM 1H-[1,2,4]oxadiazolo[4,3-a]quinoxalin-1-one (an inhibitor of soluble guanylate cyclase; 16+/-4 vs 1+/-1% and 44+/-7 vs 7+/-1%; n=10; P<0.05). 5. Thus, dilator responses of the rat basilar artery to NO appear to be mediated by activation of soluble guanylate cyclase and partially by activation of a 4-aminopyridine-sensitive mechanism. The most likely mechanism would appear to be activation of voltage-dependent K+ channels by NO/cyclic GMP.