Values obtained for cardiac output (CO) were compared using thermodilution (TD) with those obtained using bioimpedance (Bi) as measured using the Bomed NCCOM3 (Revision 6) in 28 consecutive patients in the first 24 h after coronary artery bypass surgery (CABS). In 46 paired measurements made in the first 12 h after CABS Bi values for CO were significantly lower than TD values, the limits of agreement between the two methods were also unacceptably large (mean Bi 4.38 (SD 1.40) l/min, mean TD 5.46 (SD 1.19) l/min, limits of agreement -3.05 to +0.89). In 55 paired measurements made after 12 h (all in spontaneously breathing patients) there was no significant difference between the two methods and acceptable limits of agreement, mean Bi 5.69 (SD 1.2) l/min mean TD 5.6 (SD 1.2) l/min, limits of agreement -0.99 to +1.17). The significantly lower BiCO values obtained in the first 12 h after CABS show that BiCO measurement is not consistently reliable in the intensive care setting.