Analysis of heart rate variability (HRV) by Holter monitoring is the method of choice for assessing the cardiac autonomic regulation. Rapid ventricular stimulation also provokes changes in the autonomic nervous system tone. The aim of this study was to compare time and frequency domain analysis of HRV (Elatec version 3.02) with variations of HR observed after incremental ventricular stimulation to 200/min in 130 patients. In 80 patients, ventricular stimulation provoked an initial acceleration in HR followed by a slowing with a variation of over 10%. In the other 50 patients, these variations were not observed. Holter analysis of HRV showed concordance between the two methods. In time domain analysis, the standard deviation of normal R-R intervals, the coefficient of variability [(CV = SD/mean RR) and percentage of adjacent RR intervals with a difference of more than 50 msec (pNN 50) were significantly reduced in the abnormal group, the respective values in the normal and abnormal groups being: SD 122 vs 72 msec; CV 15 versus 9% and pNN50 9 versus 5%)]. In frequency domain analysis, there was a reduction of low and high frequency spectra and of the ratio of low/high frequencies in abnormal subjects. The authors conclude that the disappearance of HR changes after ventricular stimulation is correlated to the absence of HRV on Holter recording. This simple test may be performed systematically during electrophysiological investigations. The measurements are reproductible and the results are not affected by arrhythmias or technical problems of quality recording by the Holter method which may affect analysis of HRV.