We reviewed the differential validation methods of nursing diagnoses and retrieved a total of six differential validation methods, which are Gordon & Sweeney model, Fehring model, Hoskins model, latent class model, multiple correspondence analysis and decision tree. Through the comparison and analysis of the main contents, characteristics and application status of the differential validation methods, we summarized the advantages and limitations of various methods, and future research directions, in order to provide reference for clinical nursing staff to carry out the differential validation research of nursing diagnoses.