Objective: To establish a strategy combining specific erythrocyte agglutination screening with molecular modeling for the efficient preparation of anti-M blood group antibodies, and to verify their clinical detection efficacy. Methods: An agglutination screening strategy targeting intact erythrocytes was adopted, combined with flow cytometry sorting to obtain anti-M specific B cells for BCR sequencing. Evolutionary biological analysis was performed on the antibody variable region sequences, and molecular modeling was integrated to predict antigen-antibody binding affinity. Candidate sequences were screened to construct expression plasmids, and recombinant antibodies were expressed in a eukaryotic system. Erythrocyte agglutination test, titer determination and flow cytometry were used to verify antibody binding specificity, and erythrocyte samples with different MNS phenotypes were employed to evaluate its blood group identification efficacy. Results: From the blood sample with an anti-M antibody titer of 1:512, 96 anti-M specific B cells were obtained via specific agglutination screening, and 28 pairs of heavy and light chain variable region sequences were acquired through BCR sequencing. Six candidate antibodies were screened for expression and verification based on evolutionary biological analysis and molecular modeling, among which 2G9, 2F11 and 2H5 could specifically agglutinate M antigen-positive erythrocytes, with 2G9 showing the highest agglutination titer (1:256). Flow cytometry confirmed that 2G9 could specifically recognize the M antigen, and this antibody achieved a 100% detection rate and 0% false positive rate for M blood group in erythrocytes with different MNS phenotypes. Conclusion: The specific erythrocyte agglutination screening strategy provides a new method for the enrichment of rare blood group-specific B cells; the combination of molecular modeling and evolutionary biological analysis enables rational optimization of candidate antibodies; the anti-M blood group antibody 2G9 exhibits excellent clinical detection efficacy and can be applied for clinical M blood group identification.