For evaluating the degree of accuracy of the STI, a study on 120 healthy young male volunteers was performed. Under standardized conditions ecg, phonocardiogram and carotid arterial pulse were recorded. The following variables were measured: heart cycle interval (RRI), electrical systole (QT), electromechanical systole (QS2), and left ventricular ejection time (LVET). The arithmetic mean was utilized for the consecutive beats 1--5, 1--10, 1--15, and 1--20. The mean of the beats 1--20 was assumed as reference. The increase of the analysed heart cycles from 5 to 20 lead to a decrease of the standard error of RRI, which raised the degree of accuracy. The standard error for RRI decreased from 27.9 to 9.9 ms. Already the analysis of 5 consecutive beats for QT, QS2, and LVET reached the line of identity of the correlation slope compared to 20 analysed heart cycles. The standard error was very small (2.0; 2.9; and 3.0 ms). For clinical-pharmacological studies, the evaluation of 5 consecutive heart cycles for measuring STI leads to a very high degree of accuracy. For an exact measuring of RRI, it is necessary to analyse at least 20 heart cycles.