Oxygen radical scavengers have been shown to prevent the development of ischemic preconditioning, suggesting that reactive oxygen species (ROS) might be involved in this phenomenon. In the present study, we have investigated whether direct exposure to ROS produced by photoactivated Rose Bengal (RB) could mimic the protective effects of ischemic preconditioning, Methods In vitro generation of ROS from photoactivated RE in a physiological buffer was first characterised by ESR spectroscopy in the presence of 2,2,6,6-tetramethyl-1-piperidone (oxoTEMP) or 5,5-dimethyl-1-pyrroline-N-oxide (DMPO). In a second part of the study, isolated rat hearts were exposed for 2.5 min to photoactivated RE. After 5 min washout, hearts underwent 30 min no-flow normothermic ischemia followed by 30 min of reperfusion. Results and Conclusions The production of singlet oxygen (O-1(2)) by photoactivated RE in the perfusion medium was evidenced by the ESR detection of the nitroxyl radical oxoTEMPO. Histidine completely inhibited oxoTEMPO formation. In addition, the use of DMPO has indicated that (i) superoxide anions (O-2(.-)) are produced directly and (ii) hydroxyl radicals (HO.) are formed indirectly from the successive O-2(.-) dismutation and the Fenton reaction. In the perfusion experiments, myocardial post-ischemic recovery was dramatically impaired in hearts previously exposed to the ROS produced by RE photoactivation (O-1(2), O-2(.-), H2O2 and HO.) as well as when O-1(2) was removed by histidine (50 mM) addition. However, functional recovery was significantly improved when hearts were exposed to photoactivated RE in presence of superoxide dismutase (10(5) IU/L) and catalase (10(6) IU/L). Further studies are now required to determine whether the cardioprotective effects of Rose Bengal in presence of O-2(.-) and H2O2 scavengers are due to singlet oxygen or to other species produced by Rose Bengal degradation.