Ventricular arrhythmias present with strongly varying intensity. This spontaneous variability makes it difficult to use one of the existing arrhythmia grading systems for risk or therapy efficacy studies. We attempted to explain the variability by the changing autonomic, ischaemic, circadian, and rhythmic factors. Four (two learning, two test) 24-h Holter tapes were made within one month in 31 patients with chronic frequent ventricular ectopic beats of miscellaneous aetiology and under constant drug regimen. The data were segmented into 5-min episodes, in which ectopy (dependent variable) was measured, together with heart rate, amount of heart rate variability, fraction low-frequency heart rate variability, ST depression, and clock time (independent variables). Forty-three percent of the fluctuations in arrhythmia incidence could be explained in 36% of the cases. Our study demonstrates that much of the spontaneous variability of ventricular arrhythmias can be attributed to the varying conditions. This methods of dealing with arrhythmia variability might lead to an alternative to the current arrhythmia grading systems used in risk and drug efficacy studies.