Background: Growing evidence indicates that periodic limb movements of sleep (PLMS) may be related to increased risk of developing cardiovascular disease. However, the association of PLMS with atrial fibrillation (AF) is unclear, especially in patients with sleep-disordered breathing (SDB). This study sought to investigate whether PLMS were associated with increased AF prevalence, independent of established risk factors. Methods: We performed a cross-sectional study of patients who underwent attended polysomnography at Mayo Clinic from 2011 to 2014. The association of PLMS with AF prevalence was estimated by using logistic regression models. Results: 15,414 patients were studied, 76.3% of individuals with SDB defined by apnea-hypopnea index (AHI) >= 5/h, and 15.3% with a diagnosis of AF. In univariate logistic modelling, individuals with periodic limb movement index (PLMI) >= 30/h had higher odds of AF (odds ratio [OR] 1.96, 95% confidence interval [CI]1.79-2.16, p < 0.001) when compared to patients with PLMI < 15/h. After multivariate adjustment (for age, race, sex, history of smoking, hypertension, diabetes, coronary artery disease, heart failure, cerebrovascular disease, renal disease, iron deficiency anemia, chronic obstructive pulmonary disease, AHI, arousal index), in mild SDB patients, a PLMI >= 30/h or periodic limb movement arousal index (PLMAI) >= 5/h had significantly higher odds of AF than those with PLMI < 15/h (OR 1.21, 95% CI 1.00-1.47, p = 0.048) or PLMAI < 1/h (OR 1.27, 95% CI 1.03-1.56, p = 0.024). Conclusions: Frequent PLMS are independently associated with AF prevalence in patients with mild SDB. Further studies are needed to better understand the relationship with incident AF. (C) 2017 Elsevier B.V. All rights reserved.