Cardiac adverse effects are among the leading causes of the discontinuation of clinical trials and the withdrawal of drugs from the market. The novel concept of ‘hidden cardiotoxicity’ is defined as cardiotoxicity of a drug that manifests in the diseased (e.g., ischemic/reperfused), but not in the healthy heart or as a drug-induced deterioration of cardiac stress adaptation (e.g., ischemic conditioning). Here, we aimed to test if the cardiotoxicity of a selective COX-2 inhibitor rofecoxib that was revealed during its clinical use, i.e., increased occurrence of proarrhythmic and thrombotic events, could have been revealed in early phases of drug development by using preclinical models of ischemia/reperfusion (I/R) injury. Rats that were treated with rofecoxib or vehicle for four weeks were subjected to 30 min. coronary artery occlusion and 120 min. reperfusion with or without cardioprotection that is induced by ischemic preconditioning (IPC). Rofecoxib increased overall the arrhythmias including ventricular fibrillation (VF) during I/R. The proarrhythmic effect of rofecoxib during I/R was not observed in the IPC group. Rofecoxib prolonged the action potential duration (APD) in isolated papillary muscles, which was not seen in the simulated IPC group. Interestingly, while showing hidden cardiotoxicity manifested as a proarrhythmic effect during I/R, rofecoxib decreased the infarct size and increased the survival of adult rat cardiac myocytes that were subjected to simulated I/R injury. This is the first demonstration that rofecoxib increased acute mortality due to its proarrhythmic effect via increased APD during I/R. Rofecoxib did not interfere with the cardiprotective effect of IPC; moreover, IPC was able to protect against rofecoxib-induced hidden cardiotoxicity. These results show that cardiac safety testing with simple preclinical models of I/R injury uncovers hidden cardiotoxicity of rofecoxib and might reveal the hidden cardiotoxicity of other drugs.