Kawasaki disease (KD) is a systemic vasculitis syndrome that leads to coronary artery aneurysm (CAA). While echocardiography is the most important imaging modality for coronary artery assessment, a specific diagnostic biomarker complementary for CAA has not been reported. We aimed to analyze the profiles of exosomal miRNAs extracted from the serum of KD patients and controls to identify candidate biomarkers for CAA. Serum samples from 39 healthy children, 42 CAA patients, 38 coronary artery dilatation (CAD) patients and 45 virus‐infected patients including 24 EBV patients and 21 ADV patients were randomly selected. Next generation sequencing was used to analyze serum exosomal miRNA to detect differentially expressed miRNAs. Biomarker candidates were validated by qRT‐PCR. One hundred (and) ninety‐six differentially expressed miRNAs (DEMs) were detected in CAA patients and healthy children. There were 70 DEMs and 140 DEMs in CAA patients versus CAD patients, and in CAA patients versus virus‐infected patients, respectively. We selected the three most upregulated (let‐7i‐3p, miR‐17‐3p, and miR‐210‐5p) and the three most downregulated miRNAs (miR‐6743‐5p, miR‐1246, and miR‐6834‐5p) in the DEMs, which were expressed differentially in CAA patients versus healthy children, and in CAA patients versus virus‐infected patients, not in virus‐infected patients versus healthy children, as biomarker candidates. Excluded DEMs of CAD and virus‐infected patients, let‐7i‐3p was detected by sequence data analysis as a biomarker candidate for CAA patients, and then validated by qRT‐PCR in a larger set of clinical samples. As a biomarker candidate, let‐7i‐3p provides an additional means of diagnosing CAA patients. Additionally, miRNA biomarkers complement ultrasonic imaging, allowing for greater diagnostic precision. © 2019 IUBMB Life, 2019