Objective: We investigated whether ischemic preconditioning (PC) may modify mitochondrial permeability transition (MPT) pore opening. Methods: In protocol 1, New Zealand White rabbits underwent either no intervention (sham group) or 10 min of ischemia followed by 5 min of reperfusion, preceded (PC) or not (C; control) by one episode of 5 min of ischemia and 5 min of reperfusion. Rabbits were pretreated. by either saline or the MPT pore inhibitor cyclosporin A (CsA), or its non-immunosuppressive derivative Cs29 (10 mg/kg, IV bolus). Hearts were harvested and mitochondria isolated for further assessment of Ca2+-induced MPT using a Ca2+-sensitive micro-electrode. In protocol 2, C and PC hearts underwent 30 min of ischemia and 4 h of reperfusion. They were pretreated either by saline, CsA or Cs29, as in protocol 1. Infarct size was assessed by triphenyltetrazolium, and apoptosis by TUNEL staining. Results: In protocol 1, the Ca2+ overload required to induce MPT pore opening was significantly higher in PC than in C hearts. CsA and Cs29 significantly increased the Ca2+ overload required for MPT pore opening. In protocol 2, mean infarct size averaged 25% of the risk region in CsA/Cs29 treated hearts versus 15% in PC and 55% in controls (P<0.05 vs. C, P = ns vs. PC). Cardiomyocyte apoptosis was significantly reduced by PC and cyclosporin treatment with a mean apoptotic index of less than 2% in either group versus more than 11% in controls. Conclusion: This suggests that delayed opening of MPT pore may play a major role in ischemic PC.