The topic of this work is non-invasive measuring of cardiac output based on its determining by electrocardiogram measurement. The main physiological parameters of cardiac output change and their influence on measured voltage of heart dipole are analysed. The basic cardiac output factors, which cause the change of measured heart voltage, are determined. The model of half triple axis ellipsoid has presented the formation of heart dipole. The fundamental influences, which cover the impact of size and cardiac output change on measured heart voltage, especially respiration, which causes dynamical changes of heart position and dynamical changes of resistance, have been measured. The impact of respiration is suppressed by the introduction of normalization. It has been presupposed that the impact of other artefacts would stay unchanged during measurement, and could be subsequently taken into consideration through the calibration sequence of the reference counts of the measured R wave amplitude. The applied normalization and the calibration sequence are valid only for the measured body position (for example, the upright position of the body). In the beginning, i.e. in the first learning phase, the application of the new method induces the change of the cardiac output by changing the load, from the state of inactivity to the maximum load, and, by using some of the non-invasive methods measurement the cardiac output, the mean value of respiration, the R peak amplitude and the heart rhythm are measured in the same points, generating the calibration sequence. In the second phase, the measurement phase, only the mean values of respiration, R wave amplitude and the heart rhythm are measured (every 30 seconds, because the Cardiac Output does not significantly change within that period), the result of which is a sequence of comparable items. These items are then compared with the calibration sequence. By using the method of comparison, readings of the cardiac output are deduced for each period of the calculation of mean value. The described method provides non-invasive and simple measurement of cardiac output, with simultaneous measurement of the ECG and respiration. This method is particularly suitable for long-term measurements, follow-up of patient status, ergonometric test measurements, etc.