Because central venous O2 saturation (superior vena cava, ScvO2) can be monitored with less risk of the patients than mixed venous O2 saturation (pulmonary artery, SvO2), there have been studies to see if ScvO2 could replace SvO2. But previous studies showed that these two measurements were correlated but "not interchangeable. Therefore the authors compared right atrial C#b saturation (SraO2) with S vO2 over a wide range of cardiorespiratory status including control, hemorrhage, resuscitation, and hypoxia in anesthetized cats. We performed thoracotomy and inserted cannulae directly into the right atrium and the pulmonary artery. Blood sampling were obtained synchronously through the cannulae and tested immediately. The correlation coefficients in control, hemorrhage, resuscitation, and hypoxia groups were 0.876, 0.794, 0.946, 0.948 respectively and the two measurements in each group showed statistically significant correlations (p<0.05). But the biases of the two measurements in each group were 0.11±2.9, 0.35±4.2, -0.55±3.2, 0.23±4.2 respectively and the limits of agreement ( 2 standard deviation) in all groups exceeded permissible (5%) to conclude that the two measurements were in agreement. Thus, we reached the conclusion that the absolute values of SraO2, though not being sufficiently identical to S vO2 to calculate O2 uptake or pulmonary shunt precisely, can reflect the S vO2 trend following the O2 supply / demand change. Further clinical studies are needed.