The purpose of this study was to establish a standardized protocol in order to obtain accurate and reproducible cardiac output (CO) measurements by the thermodilution technique in the unsedated calf. In 5 healthy calves, the effect of various temperatures, volumes, rates, means and sites of injection of the thermal indicator were tested for their effect on accuracy and reproducibility of CO measurement by the thermodilution technique, the Fick method being used as the standard technique. Five milliliters of an iced 5% dextrose solution per 100 kg body weight constituted the correct amount of thermal indicator. Manual injection of the indicator bolus and/or injection of this bolus through the proximal lumen of a thermodilution Swan-Ganz catheter gave poor results of CO measurements by the thermodilution technique. On the other hand, results of CO measurements were highly reliable when the bolus injection was performed by an ECG-synchronized injection pump and/ or through a right atrial catheter with a larger diameter and a shorter length than the classical Swan-Ganz catheter. It was concluded that an iced 5% solution of dextrose (5 ml/100 kg body weight) injected through a right atrial catheter by an ECG-synchronized pump gives the most accurate and reproducible CO measurements by the thermodilution technique in unsedated calves.