The role of ATP-sensitive potassium channels (KATP) in the mechanism of ischemic preconditioning (PC) is controversial, partly because descriptions of inhibition of PC by KATP blockers in the literature are inconsistent. We sought a reason for the discrepant findings regarding the effects of glibenclamide (Glib), a specific blocker of KATP, in preventing the reduction of infarct size (IS) induced by PC. The effect of Glib pretreatment (0.3 mg/kg i.v.) on PC was examined in three conditions: (a) when PC was performed with 3- and 5-min ischemia (i.e., potency of PC differs), (b) when rabbits were pretreated with prazosin and metoprolol (0.15 mg/kg i.v. each) to reduce myocardial O2 consumption, and (c) when xylazine was added to pentobarbital anesthesia. In rabbits under pentobarbital anesthesia, the left coronary artery was occluded for 30 min and then reperfused. The area at risk (AAR) and IS were determined 72 h after reperfusion in the first series of experiments and 3 h after reperfusion in the second and third series. IS as a percentage of AAR (%IS/AR) were 31.7 +/- 2.8 and 19.6 +/- 2.5% (SEM) after PC with 3- and 5-min ischemia, respectively, values significantly smaller than %IS/AR in the untreated control group (49.2 +/- 3.3%). The limitation of IS observed with 3- or 5-min PC was not prevented by Glib. Glib also failed to block %IS/AR reduction by PC, even when rate-pressure product (RPP) was reduced to approximately 65% by prazosin/metoprolol (Praz/Met) pretreatment.(ABSTRACT TRUNCATED AT 250 WORDS)