Abstract Background Atrial fibrosis is a major determinant of atrial fibrillation (AF) susceptibility and ablation targeting fibrotic areas to suppress AF revealed by magnetic resonance imaging (MRI) was considered a promising approach to treat persistent AF; however, its outcome has been disappointing. We sought to develop a new imaging modality to detect the severity and distribution of atrial fibrosis. Methods This study enrolled 19 healthy volunteers and 85 patients with various cardiovascular (CV) diseases of whom 22 had persistent AF. Each patient underwent 18F-labeled aluminum fluoride targeting fibroblast activation protein inhibitor (18F-AlF-FAPI)-positron emission tomography - magnetic resonance imaging (PET-MRI). Left atrial appendage tissue biopsied at 18F-FAPI-uptake positive and negative sites was stained with Masson trichome to detect fibrosis (N=10). We compared cardiac atrial 18F-AlF-FAPI-PET-MRI imaging among healthy individuals, patients without AF and patients with persistent AF as well as patients developing and not developing postoperative AF (POAF). Results Degree of fibrosis based on Masson staining was consistent with 18F-FAPI-uptake (P<0.05). Whereas 18F-FAPI uptake was essentially undetectable in healthy volunteers, in patients with CV diseases, 18F-FAPI uptake was present in all atrial chambers. Patients with persistent AF had the greatest and most widespread uptake (P<0.001), including left atrium (P=0.001), right atrium (P<0.001), left atrial appendage (P=0.007), right atrial appendage (P<0.001), and left atrial posterior wall (P=0.004). 18F-FAPI-uptake did not differ between those with versus without POAF. Conclusions 18F-AlF-FAPI-PET-MRI imaging detected atrial fibrosis in patients with various CV diseases; patients with persistent AF exhibit severe and widespread fibrosis, while POAF is not associated with increased 18F-FAPI uptake.