The transcutaneous partial pressure of oxygen (PtcO₂) index has been used to detect low-flow state in circulatory failure, but the value of the transcutaneous oxygen challenge test (OCT) to estimate low cardiac output has not been thoroughly evaluated. The prospective observational study examined 62 septic patients requiring PiCCO-Plus for cardiac output monitoring. Simultaneous basal blood gases from the arterial, central venous catheters were obtained. Cardiac indices were measured by the transpulmonary thermodilution technique at the same time, then the 10-min inspired 1.0 fractional inspired oxygen concentration (FIO₂) defined as the OCT was performed. Transcutaneous partial pressure of oxygen was measured continuously by using a noninvasive transcutaneous monitor throughout the test. The values for arterial pressure of oxygen (PaO₂) were examined on inspired of 1.0 FIO₂. We calculated the PtcO₂ index = (baseline PtcO₂/baseline PaO₂), 10-min OCT (10 OCT) = (PtcO₂ after 10 min on inspired 1.0 O₂) - (baseline PtcO₂), and the oxygen challenge index = (10 OCT) / (PaO₂ on inspired 1.0 O₂ - baseline PaO₂). Patients were divided into two groups: a normal cardiac index (CI) group with CI of greater than 3 L/min per m (n = 41) and a low CI group with CI of 3 L/min per m or less (n = 21). The 10 OCT and the oxygen challenge index predicted a low CI (≤ 3 L/min per m) with an accuracy that was similar to central venous oxygen saturation, which was significantly better than the PtcO₂ index. For a 10 OCT value of 53 mmHg, sensitivity was 0.83; specificity, 0.86; a positive predictive value, 0.92; and a negative predictive value, 0.72 for detecting CI of 3 L/min per m or less. We propose that the OCT substituted for the PtcO₂ index as an accurate alternative method of PtcO₂ for revealing low CI in septic patients.