We evaluated the accuracy of fiberoptic catheter oximetry in the jugular bulb during conditions of normothermia, hemodilution, and hypothermia in 11 patients who underwent cardiac surgery with cardiopulmonary bypass (CPB). An oximetry catheter was inserted into the right jugular bulb under general anesthesia, calibrated by the in vitro (n = 7) or in vivo (n = 4) mode. Jugular bulb oxygen saturation (SjO2) with the catheter oximeter was compared with a concurrent laboratory CO-oximeter value from a blood sample during surgery. Nasopharyngeal temperature (NPT) and hemoglobin concentration (Hb) were also measured. The oximetric catheter SjO2 correlated closely with the CO-oximeter determinations in both calibration modes (in vitro, r2 = 0.88; in vivo; r2 = 0.96). Data in the in vitro calibrated group were grouped into three conditions; 1) normothermia and no hemodilution, 2) normothermia and hemodilution, and 3) hypothermia and hemodilution, and showed good correlations between SjO2 values measured by the two methods (r2 = 0.90, r2 = 0.81, r2 = 0.79, respectively). The difference in SjO2 values by the two methods was not significantly affected by changes in NPT and Hb during CPB. In conclusion, the continuous SjO2 monitoring with catheter oximetry during CPB would be accurate and reliable under either calibration mode. Moderate hypothermia and hemodilution during CPB did not significantly influence the accuracy.