Permanent electrostimulation is the therapy of choice for syncope due to bradycardic rhythm disturbance. In maintaining optimal hemodynamic function, the role of atrial systole has been well recorgnized and the time relationship between atrial and ventricular systole have great relation with cardiac output. Assessment of optimal mode of cardiac pacing by nonivasive method is very important for the best hemodynamic effect. To evaluate the usefulness of systolic time interval for optimal pacing mode, we calculated systolic time interval by using ventricular pacing spike on electrocardiogram to aortic opening time/left ventricular ejection time by aortic pressure curve (invasive PEP/LVET) in various modes of cardiac pacing and measured cardiac output by thermodilution method simultaneously in 9 mongrel dogs. Basal pacing cycle length were 300 msec, and the atrioventricular (AV) and ventriculoatrial (VA) interval during AB & VA sequential pacing were set at 30 msec, 60 msec and 90 msec. The result were as follows: 1) The cardiac output at AV interval of 90 msec (1.65+/-0.23 L/min) is significantly higher than 30 msec (1.38+/-0.19 L/min) in AV sequential pacing. 2) The cardiac output in ventricular pacing is higher than VA sequential pacing, but no significant changes noted among VA interval 90 msec, 60 msec and 30 msec. 3) The invasive PEP/LVET at VA interval of 90 msec (0.85+/-0.17) is significantly lower than 60 msec (0.97+/-0.16) and 30 msec (1.01+/-0.16) in AV sequential pacing. 4) The invasive PEP/LVET among VA sequential pacing with 90 msec, 60 msec, 30 msec interval and ventricular pacing did not show any significant difference. 5) When AV interval changes from 90 msec to 30 msec during AV sequential pacing, cardiac output decreased and invasive PEP/LVET increased. 6) In VA sequential pacing, there were no changes of cardiac output and invasive PEP/LVET when VA interval changes from 90 msec to 30 msec. In conclusion, systolic time interval can be used for estimation of hemodynamic changes during AV sequential pacing considering our results and other authors' results of high correlation between invasive & noninvasive PEP/LVET.