1. In conscious rats, alpha-trinositol (D-myo-inositol-1, 2, 6 triphosphate; 5-80 mg kg-1 h-1 infusion) caused dose-dependent hypotension, tachycardia and hyperaemic dilatation in renal, mesenteric and hindquarters vascular beds. These effects were accompanied by inhibition of the renal vasodilator effects of acetylcholine (ACh), and of the mesenteric vasodilator effects of sodium nitroprusside (SNP) and, particularly, of levcromakalim (LCK). 2. In the light of the latter finding, in a second experiment, we assessed the influence of the KATP channel inhibitor, glibenclamide (20 mg kg-1), on resting haemodynamics, on responses to ACh, bradykinin (BK), SNP and LCK, on the haemodynamic action of alpha-trinositol, and on the effects of the latter on responses to the vasodilators, over a period of 3 days. 3. In the presence of saline, glibenclamide caused a reproducible pressor effect, accompanied by renal, mesenteric, and hindquarters vasoconstrictions on all 3 experimental days; these effects were unrelated to changes in blood glucose. In the presence of glibenclamide, only the hindquarters vasodilator response to BK, and all the cardiovascular actions of LCK were inhibited. 4. On the first experimental day, the hindquarters vasodilator effect of alpha-trinositol was substantially inhibited by glibenclamide, the renal vasodilatation less so, and the mesenteric vasodilatation not at all. However, over the subsequent two days, the mesenteric vasodilator effect of alpha-trinositol became increasingly sensitive to glibenclamide. 5. In the presence of alpha-trinositol and glibenclamide, on the first experimental day, the inhibition of the renal vasodilator effect of ACh was no greater than with alpha-trinositol alone in the first experiment. Moreover, on the third experimental day, both before and after glibenclamide, the inhibition by alpha-trinositol of the renal vasodilator response to ACh was less than on the first experimental day. Similarly, the alpha-trinositol-induced inhibition of the mesenteric vasodilator effect of SNP, and of the hindquarters vasodilator action of BK, waned over the 3 experimental days. The inhibition of the cardiovascular effects of LCK were similar on all 3 experimental days, but no greater in the presence of alpha-trinositol and glibenclamide than with glibenclamide alone. 6. These results indicate that KATP channels are involved in the maintenance of resting vasodilator tone in renal, mesenteric and hindquarters vascular beds. However, although additional activation of KATP channels is responsible for all the haemodynamic effects of LCK, it contributes only to the hindquarters vasodilator action of BK and is not involved in any of the responses to ACh or SNP. The hindquarters, renal and mesenteric vasodilator effects of alpha-trinositol may involve (in the same rank order) activation of KATP channels, probably through an indirect mechanism. However, it is unlikely that direct or indirect interaction of alpha-trinositol with KATP channels explains the ability of the drug to inhibit the renal vasodilator action of ACh, or the mesenteric vasodilator effects of SNP or LCK.