Accurate measurement of core temperature is of utmost importance during on-pump cardiac surgery, for detection of hypothermia before cardiopulmonary bypass (CPB), guidance of temperature management on CPB, active rewarming on CPB and guidance of warming therapy after CPB. Most temperature measurement methods are known to become inaccurate during rapid changes in core temperature and suffer from delayed detection of temperature changes. Zero-heat-flux temperature (ZHF) measurement from the lateral forehead may be an alternative, non-invasive method quantifying the core temperature. A prospective, observational, multicentre study was conducted in one hundred patients scheduled for on-pump coronary artery bypass grafting. Core temperatures were measured every minute by two zero-heat-flux thermometer (SpotOn™) and a bladder thermometer and a pulmonary artery catheter (PAC) in the period after induction of anesthesia until CPB. Accuracy and precision of both methods were compared against core temperature measured in the pulmonary artery using the method of Bland and Altman. A high accuracy (around 0.1 °C) and a very good precision (Limits of agreement (LoA) − 0.6; 0.4 °C) were found between zero-heat-flux thermometer and core temperature measured by PAC. Among the two ZHF thermometers the bias was negligible (− 0.003 °C) with narrow LoA of − 0.42 °C and 0.41 °C. In contrast, bias between bladder temperature and PAC temperature was large (0.51 °C) with corresponding LoA of − 0.06 °C and 1.1 °C. ZHF thermometers are in contrast to bladder temperature a reliable core temperature monitor in cardiac surgery during the period after induction of anestesia until CPB. The zero-heat-flux method can provide clinicians reliably with continuous and non-invasive measurements of core temperature in normothermic and mild hypothermic temperature ranges and therefore can be helpful to guide temperature management.