BACKGROUND We aimed to evaluate whether clinical factors, which influence heart rate variability (HRV) in the presence of undisturbed sinus rhythm, have any associations with HRV in patients with permanent atrial fibrillation (AF). METHODS One hundred ninety-seven consecutive patients with permanent AF were included (122 males, 75 females, aged 64 ± 11 years, range 25-85). In each patient a 24-hour electrocardiographic recording was performed and an HRV fraction (HRVF)-the index based on scatter plot numerical processing-was calculated. Additionally, standard HRV measures were analyzed. Reduced HRVF was defined as its value lower than lower normal limit. Demographic and clinical factors were examined for their association with a reduced HRVF by means of a univariate and multivariate logistic regression analysis. RESULTS The reduced HRVF was associated with advanced age, clinical diagnosis of a previous MI or dilated cardiomyopathy, presence of diabetes, depressed left ventricular function, NYHA class > II, treatment regimen, use of digoxin, diuretics or antiarrhythmic agents, nonuse of beta-blockers, and increased heart rate. The independent determinants that sustained after multivariate analysis were: heart rate (per 10 bpm increase, odds ratio 2.77 [1.88-4.07]), age (per 5 years increase 1.43 [1.1-1.85]), depressed left ventricular EF (<30% vs higher 2.26 [1.19-4.31]), and presence of diabetes (3.45 [1.1-10.85]). The HRVF correlated moderately with standard HRV measures. This index showed also the strongest correlation with left ventricular ejection fraction. CONCLUSIONS We concluded that advanced age, left ventricular systolic dysfunction, increased heart rate, and presence of diabetes are cofactors of a reduced HRV in AF patients. Thus, the determinants of heart rate variability in the presence of atrial fibrillation are the same as those in sinus rhythm.