We investigated whether endogenous ligands of peroxisome proliferator-activated receptor-gamma (PPAR-gamma) protect the heart against ischemia-reperfusion (I/R) injury. The selective PPAR-gamma antagonist GW9662 (2-chloro-5-nitrobenzanilide) was used in rat models of 1) regional myocardial I/R, 2) ischemic preconditioning, and 3) delayed cardioprotection by endotoxin. We also investigated the effects of the selective cyclooxygenase-2 inhibitor, parecoxib, on ischemic preconditioning and delayed cardioprotective effects of endotoxin. Male Wistar rats were anesthetized with sodium thiopentone. Animals were subjected to either 15 or 25 min of regional myocardial I/R and pretreated with the PPAR-gamma agonist ciglitazone (0.3 mg/kg), the PPAR-gamma antagonist GW9662 (1 mg/kg), or GW9662 and ciglitazone. Animals were also subjected to either 1) ischemic preconditioning alone, ischemic preconditioning, and pretreated with either GW9662 or parecoxib (20 mg/kg) or 2) lipopolysaccharide (LPS) (1 mg/kg) alone, LPS, and pretreated with ciglitazone, GW9662, or parecoxib (20 mg/kg). Myocardial infarct size was determined by p-nitroblue tetrazolium staining. The PPAR-gamma antagonist GW9662 (1 mg/kg) abolished the cardioprotection afforded by the potent PPAR-gamma agonist ciglitazone (0.3 mg/kg). Neither GW9662 nor parecoxib affected the cardioprotective effects of ischemic preconditioning. Pretreatment with ciglitazone did not provide additional cardioprotection to LPS-treated animals. Both GW9662 and parecoxib abolished the delayed cardioprotective effects of endotoxin. Thus, we propose that 1) endogenous ligands of PPAR-gamma are being generated by myocardial ischemia in sufficient amounts to attenuate myocardial I/R injury, and 2) that cyclooxygenase-2 metabolites contribute to (or even account for) the cardioprotective effects of endotoxin (second window of protection) by acting as endogenous PPAR-gamma ligands.