Inflammatory bowel disease (IBD) diagnosis is a complex process that relies on the integration of multiple clinical assessments, including endoscopy and histopathology. However, the absence of highly accurate molecular diagnostic markers remains a critical unmet need. So, we focused on the B-cell receptor (BCR), and constructed BCR repertoires database for ulcerative colitis (UC) and Crohns disease (CD) using RNA-seq data. We developed novel BCR scores for diagnosing IBD and explored their potential for disease monitoring and predicting treatment response. In multiple independent validation cohorts, the BCR score achieved average AUCs of 0.867 and 0.762 for UC and CD diagnosis, respectively. Notably, we also tested in fecal wash samples collected during colonoscopy, where the BCR score yielded AUCs of 0.791 for UC and 0.875 for CD, highlighting its potential as a high-performance, non-invasive diagnostic approach for IBD. Our study demonstrates the transformative potential of BCR-based diagnostics in improving IBD management.