To evaluate and quantify the relationship between premature ventricular contractions (PVCs) and heart rate (HR), 57 patients (48 men and 8 women, mean age 59.8 +/- 7.9 years) with severe PVCs (Lown-Wolf grade > or = 3a) over 24 hours of Holter monitoring were studied. Twenty had no coronary artery disease (CAD), 25 had angiographically documented CAD, and 12 had acute myocardial infarction. All parameters of the 24-hour recordings from two ECG leads were measured by a Holter analyzer designed in our laboratory, based on fast microprocessors and controlled by a microcomputer. Scatter diagrams of the number of PVCs per minute as a function of HR and correlation coefficient were computed for various HR values corresponding to a total number of minutes greater than five. A positive correlation (r > or = 0.35) was found in most patients without CAD (85%); there was a complex relationship between the strength of the correlation and the presence of CAD or acute myocardial infarction because of a greater variability in the results of correlation coefficient analysis (coefficient of variation 62%, 208%, and 145% in patients without CAD, with CAD, and with acute myocardial infarction, respectively). The incidence of a positive correlation was similar in patients with Lown-Wolf grade III (63%), IVa (82%), or > or = IVb (67%) arrhythmias. The reproducibility of the correlation coefficient of the relationship between PVC frequency and HR was tested in 15 patients. The mean value of the correlation coefficient was 0.801 +/- 0.169 for the first test and 0.805 +/- 0.22 (p = NS) for the second test.(ABSTRACT TRUNCATED AT 250 WORDS)