OBJECTIVE To assess the relative contributions of rate control and rhythm regularization to left ventricular function in atrial fibrillation (AF) patients undergoing atrioventricular nodal ablation. This was performed by assessing the effect of ventricular rhythm regularization on left ventricular function during AF, and the effect of varying heart rate on left ventricular function after ablation. METHODS Eleven patients with continuous AF and V/VI-R pacemakers undergoing therapeutic atrioventricular nodal ablation were studied. Preablation patients underwent two 30 min observation periods in a randomized, blinded fashion during which they were either in baseline AF (pacer set to default V/VI 50/min) or being paced using a rhythm stabilizing algorithm (RSA) designed to regularize rhythm without changing baseline ventricular rate. Six weeks after ablation, patients were again observed during the two following 30 min periods: pacing at a low clinically indicated rate (69+/-9 beats/min), and pacing at the rapid, mean preablation rate. During all observation periods, left ventricular function was measured continuously using a nuclear vest that provided validated measures of heart rate, ejection fraction, and normalized end-systolic volume (ESV) and end-diastolic (EDV) volume. RESULTS Before ablation, RSA successfully regularized rhythm, decreasing the coefficient of variation of interbeat intervals 20+/-5% to 10+/-4% (P<0.001). The heart rate with RSA (105+/-19 beats/min) was not significantly different from the baseline AF rate (102+/-21 beats/min). Increased rhythm regularity achieved by RSA significantly improved left ventricular function, decreasing ESV from 62+/-12 units to 57+/-11 units (P=0.03), and increasing the ejection fraction from 31+/-11% to 36+/-11% (P=0.03). After ablation, at the clinically indicated low pacing rate of 69+/-9 beats/min, a much greater improvement in ejection fraction was observed, increasing to 44+/-13% (P=0.005 compared with preablation). However, rapid regular pacing at the mean preablation rate of 110+/-18 beats/min eradicated this improvement, decreasing the ejection fraction to 31+/-8% (P=0.003), and increasing ESV from 53+/-13 units to 62+/-8 units (P=0.006). CONCLUSIONS Rhythm regularity achieved by a regularizing pacing algorithm can significantly, albeit modestly, improve left ventricular function in AF. However, more marked improvements in left ventricular function seen after ablation are primarily due to rate reduction alone.