BACKGROUND AND OBJECTIVES: Cardiac performance is dependent on both the preceding RR interval (RR-1) and the prepreceding RR interval (RR-2) in atrial fibrillation (AF). We reported a new method for improving the relation between the two RR intervals and cardiac performance in AF of various causes. The aim of this study was to re-evaluate the method and its relationship in AF without organic heart disease. SUBJECTS AND METHODS: The beat to beat variation in the left ventricular outflow peak ejection velocity (Vpe) was measured by pulsed Doppler ultrasound in 28 consecutive patients with lone AF. The relations between the RR-2 and the Vpe were obtained before and after the exclusion of coordinates with the RR-1<0.5 second. The association of the Vpe with the RR-1 was adjusted by the RR-2 using an equation obtained from the relation between the RR-2 and the Vpe. RESULTS: The RR-2 was found to have a weak, negative, association with the Vpe. The mean squared correlation coefficient (r2) between the RR-2 and the Vpe was 0.14+/-0.13, which was improved to 0.23+/-0.21 (p=0.007) following the exclusion of coordinates with a RR-1<0.5 second. The RR-1 was positively associated with the Vpe. The mean r2 between the RR-1 and the Vpe was 0.55+/-0.15, which became stronger, at 0.68+/-0.12 (p<0.001), following adjustment with the RR-2. A multiple stepwise regression analysis revealed that the mean and standard deviation of the RR interval, and the duration of AF were independently associated with the modified r2 between the RR-2 and the Vpe. CONCLUSION: Simple modification could improve the relationship of both the RR-1 and the RR-2 with the cardiac performance in AF without organic heart disease, as with AF of various causes.