Abstract:[Background] The isolation and identification of Cronobacter forms the theoretical basis for precise treatment in clinical medicine. However, the current biochemical identification methods can only identify them at the genus level. [Objective] To develop a simple and reliable method for species identification of Cronobacter, we established a numerical identification system and evaluated its identification performance. [Methods] The biochemical reaction formulations required for the identification system were concentrated and integrated into a set of plastic containers to prepare identification strips. The numerical identification system of Cronobacter was developed based on the numerical identification theory, and the identification accuracy was assessed with the standard strains and isolates of Cronobacter. The identification results based on the sequencing of fusA were taken as the reference and compared with those obtained with the numerical identification method. [Results] We established a biochemical reaction positive probability database for distinguishing the 9 species/subspecies of Cronobacter. Simultaneously, we defined the calculation methods and result evaluation criteria for the numerical method and developed a numerical identification system. Five standard strains of Cronobacter were correctly identified by the numerical identification system, and the results were evaluated as excellent. Among them, strain DSM 18707 was accurately identified as C.dublinensis subsp. lactaridi. The total identification agreement between the numerical identification method and the fusA sequencing method of the 84 Cronobacter isolates in the test set was 100%. The numerical identification method clearly identified other strains except C.sakazakii type IV (one strain) and C.dublinensis type III (one strain) with unacceptable identification results. [Conclusion] The numerical identification system developed in this study for Cronobacter. has simple operation, low cost, and high accuracy, which can provide an experimental basis for clinical diagnosis and food safety detection and demonstrates a promising prospect of application and popularization.