This study compared the cardiac output responses to haemodynamic interventions as measured by three minimally invasive monitors (Oesophageal Doppler Monitor, the VigileoFlotrac and the LiDCOrapid) to the responses measured concurrently using thermodilution, in cardiac surgical patients. The study also assessed the precision and bias of these monitors in relation to thermodilution measurements. After a fluid bolus of at least 250 ml, the measured change in cardiac output was different among the devices, showing an increase with thermodilution in 82% of measurements, Oesophageal Doppler Monitor 68%, VigileoFlotrac 57% and LiDCOrapid 41%. When comparing the test devices to thermodilution, the kappa statistic showed at best only fair agreement, Oesophageal Doppler Monitor 0.34, LiDCOrapid 0.28 and VigileoFlotrac-0.03. After vasopressor administration, there was also significant variation in the change in cardiac output measured by the devices. Using Bland-Altman analysis, the precision of the devices in comparison to thermodilution showed minimal bias, but wide limits of agreement with percentage errors of Oesophageal Doppler Monitor 64.5%, VigileoFlotrac 47.6% and LiDCOrapid 54.2%. These findings indicate that these three devices differ in their responses, do not always provide the same information as thermodilution and should not be used interchangeably to track cardiac output changes.