n resting conditions, complex compensatory mechanisms are believed to adequately regulate blood flow to vital organs, especially to the brain. During exercise, however, the O2 demand of exercising muscles surges to levels 10-15 times higher than that in the resting condition. The blood flow to the exercising muscles increases tremendously to meet this sudden surge in O2 demand, thereby leading to a relative hypoperfusion of other organs. Our group recently studied cerebral oxygenation during exercise using near-infrared spectroscopy (NIRS) in patients with left ventricular dysfunction. 1 We found, as a result, that nearly half of the patients exhibited decreases in the cerebral oxyhemoglobin (O2Hb) during exercise, and that the decrease was more prominent in patients with a lower left ventricular ejection fraction (LVEF). 1 In these patients, cerebral O2Hb, a parameter that initially remained constant at lower work rates during incremental exercise, began to decrease at higher work rates. 1,2 The decrease in cerebral O2Hb presumably reflected a cerebral hypoxia during exercise resulting from impaired O2 transport to the brain.While risk stratification of chronic heart disease had been primarily based on resting hemodynamic measurements, cardiopulmonary exercise testing (CPX) has now become indispensable in singling out those patients with a poor prognosis. Among the parameters obtained from CPX, the peak O2 uptake (V • O2) is considered a gold standard for identifying patients with a poor prognosis and selecting candidates for cardiac transplantation. 3 The ratio of the increase in ventilation (V • E) to the increase in CO2 output (V • CO2) during exercise (V • E/V • CO2 slope) and the ratio of the increase in V• O2 to the increase in work rate (ΔV • O2/ΔWR) are also independent prognostic markers in cardiac patients. 4 In 1999, Lee et al reported that cerebral metabolism measured by magnetic resonance spectroscopy is abnormally deranged in patients with heart failure. 5 They also found that this abnormality is an independent predictor of mortality in these patients. 6 The cerebral metabolic abnormality is assumed to be due to overall and/or local disturbances of cerebral blood flow, that is, impaired cerebral oxygenation. 5 The impaired cerebral oxygenation can be detected by the non-invasive measurement of cerebral O2Hb using NIRS, and the detection might be more sensitive during exercise in which the O2 demand by exercising muscles increases tremendously, rather than measurements taken at rest. If the cerebral metabolic abnormality resulting from decreased cerebral O2 transport is related to prognosis, the decrease in cerebral O2Hb during exercise must also influence the prognosis.In the present study, we sought to establish a prognostic value of decreased cerebral O2Hb during exercise in cardiac patients. We selected patients with coronary artery disease, and compared the prognostic power of cerebral O2Hb during exercise with those of established respiratory gas indexes obtained from CPX, as cerebral oxygena...