Continuous fiberoptic measurement of mixed venous oxygen saturation (SvO2) via a pulmonary artery catheter is a useful, though invasive, monitoring technique. Continuous right atrial venous oxygen saturation measurement by oximetry offers the potential of a significantly less invasive SvO2 measurement. However, catheter motion, character of the vessel, chamber wall reflection, the filtering technique involved in calculating oxygen saturation, and the streaming of venous blood prior to ventricular mixing may influence the feasibility of continuous right atrial (RA) SvO2 measurement. This study investigated the performance of fiberoptically measured RA SvO2, at a position 2 cm from the tricuspid valve, relative to simultaneously measured pulmonary artery (PA) SvO2. Ten pigs were subjected to circulatory shock or chemically induced lung damage. Over a total monitoring period of approximately 40 hours, 464 paired data points were sampled at 5-minute intervals. The difference between the overall means of RA and PA SvO2 was 0.91% with a standard error of the estimate of 4.7%, a regression equation of RA SvO2 = PA SvO2 (0.94 + 2.1) PA So2, and a correlation coefficient of 0.94. Our conclusion, although extrapolated from a pig model, is that fiberoptic SvO2 monitoring may be accomplished less invasively and at a lower cost with a right atrial catheter.