To prevent unphysiological temperature fluctuations in the myocardium in the open-chest model, we constructed a thermocage. Five pigs under pentobarbital sodium anesthesia underwent repetitive left anterior descending (LAD) coronary artery occlusions. Myocardial temperature was measured without any thoracic temperature-controlling device and in the presence of either a heating lamp or the thermocage. Without any thoracic temperature-controlling device, the temperature at 5-mm myocardial depth was 1.28 ± 0.33°C below the intra-abdominal temperature (P < 0.05). During a proximal 5-min LAD occlusion, myocardial temperature decreased by 1.86 ± 1.02°C in the ischemic area (P < 0.05). Both the heating lamp and the thermocage abolished the difference between intra-abdominal and myocardial temperatures and prevented the decrease in myocardial temperature during ischemia. Only the thermocage minimized myocardial temperature fluctuations due to air currents and prevented epicardial exsiccation. We conclude that either a thermocage or a heating lamp may be used to normalize myocardial temperature in the open-chest pig model. However, the thermocage is superior to the lamp in minimizing temperature fluctuations and preventing epicardial exsiccation.