Objective To prepare mouse monoclonal antibodies(mAbs) against ovalbumin(OVA), and to establish and verify a double antibody sandwich ELISA quantitative detection method for the detection of OVA residues in influenza virus split vaccine.Methods Using OVA standard as immunogen, three positive hybridoma cell lines were screened by mouse hybridoma fusion technique. The OVA mAb ascites was prepared by in vivo induction method, and then purified by ammonium sulfate precipitation and Protein A affinity chromatography. The specificity of mAb was detected by Western blot, and the relative affinity was determined by indirect ELISA. Using the overlapping coefficient as the index, the optimal paired mAbs of three hybridoma cell lines were screened, and the OVA double antibody sandwich ELISA detection method was established. The concentration of coating antibody(8. 0, 4. 0, 2. 0, 1. 0, 0. 5 ??g/mL) and the dilution of detection antibody(1∶100, 1∶500, 1∶1 000, 1∶2 000) were optimized by checkerboard titration, and then the limit of quantitation(LOQ),linear range, specificity, accuracy, precision, applicability and stability of the method were verified. The OVA content of seven batches of influenza virus split vaccine bulk solutions was determined by the established method and imported OVA detection kit, and the coincidence rate was calculated.Results A total of three hybridoma cell lines that could be stably passaged and produce OVA-specific mAb were selected, named 3D6, 2B11 and 3H8, respectively, and the ascites titers were all higher than 10~7. After purification, the purity was over 95%, with good specificity, and the relative affinity ranking was2B11 > 3D6 > 3H8. After pairing screening and optimization, it was determined that 2B11 was used as the coating antibody with the optimal concentration of 4 ??g/mL, and biotin-labeled 3D6 was used as the detection antibody with the optimal dilution of 1∶100. The method had a limit of quantitation(LOQ) of 3 ng/mL. In the concentration range of 3. 13-50. 00 ng/mL,OVA standard exhibited a good linear relationship with A_(450/630), and the linear regression equation was y = 0. 004 9 x + 0. 088 7,R~2= 0. 987 9. This method had no cross-reaction with fetal bovine serum and human serum albumin. The spiked recovery rates of three influenza virus split vaccine bulk solutions were all within the range of 90% to 110%. The CVs of precision verification were less than 10%. The dilution of three batches of influenza virus split vaccine bulk solutions showed a good dose-dependent relationship with A_(450/630). Compared with the detection reagents stored at 4 ℃ for 6 days, the detection results of the detection reagents stored at 37 ℃ for 3 and 6 days decreased by less than 20%, and each R~2 was more than 0. 97.The coincidence rates between the detection results of the established method and imported OVA detection kit were between93. 96% and 100. 89%.Conclusion The OVA double antibody sandwich ELISA quantitative detection method was established by using the prepared OVA mAb, with good specificity, accuracy, precision, applicability and stability, which provides a reliable detection method for the determination of OVA residues in the verification requirements of chicken embryo-derived vaccine finished products.