Circular RNAs (circRNAs) are non-coding RNAs forming closed-loop structures, and their aberrant expression may lead to disease. However, the potential network of circRNA-associated competing endogenous RNA (ceRNA) involved in nonvalvular persistent atrial fibrillation (NPAF) has not been previously reported. In the present study, four left atrial appendages (LAA) of patients with NPAF and four normal LAAs were examined via RNA sequencing, and their potential functions were investigated via bioinformatics analysis. The circRNA-enriched genes were analyzed using Gene Ontology (GO) categories, while the enrichment of circRNAs was detected via the Kyoto Encyclopedia of Genes and Genomes (KEGG) database. A total of 296 significantly dysregulated circRNA transcripts were obtained, with 238 upregulated and 58 downregulated. A number of circRNAs were further confirmed using reverse transcription-quantitative polymerase chain reaction analysis. Furthermore, the more comprehensive circRNA-associated ceRNA networks were examined in patients with NPAF. GO categories and KEGG annotation analysis of circRNAs revealed that the circRNA-associated ceRNA networks were likely to influence AF though alterations in calcium and cardiac muscle contraction. The circRNA-associated ceRNA networks revealed that dysregulated circRNAs in NPAF may be involved in regulating hsa-microRNA (miR)-208b and hsa-miR-21. To the best of our knowledge, this study presents the circRNA-associated ceRNA networks in NPAF for the first time, which may have potential implications for the pathogenesis of AF. This study reveals a potential perspective from which to investigate circRNAs in circRNA-associated ceRNA networks (hsa_circRNA002085, hsa_circRNA001321) in NPAF, and provides a potential biomarker for AF. AF causes 130,000 mortalities/year in the USA (5) and affects 2-3% of the European population (4,6,7). The number of patients with AF is predicted to increase rapidly in the coming decades and to gradually reduce quality of life (8,9). In Asia, the number of patients with AF and AF-associated stroke is estimated to reach 72 million and 2.9 million by 2050, respectively (3,10). Warfarin is commonly applied to treat nonvalvular persistent atrial fibrillation (NPAF) and research on NPAF has been extensive (11,12), although the current treatment options remain inadequate. Thus, it is necessary to identify a novel biomarker for the prediction, diagnosis and treatment of NPAF. Circular RNAs (circRNAs), non-coding RNAs identified in 1991 (13), regulate gene expression and act as microRNA (miRNA) 'sponges' by competing with endogenous RNA (ceRNA) networks to suppress specific miRNA activity (14,15). circRNAs are associated with numerous diseases and may serve an important role in diagnosis or pathogenesis (16,17). Previous