Sepsis is a critical condition. The key factor affecting the survival of patient is whether standard treatment can be obtained timely. Because of the complexity of its pathogenesis and high heterogeneity, there is no special diagnosis method currently. Early identification is difficult. Delayed diagnosis and treatment is closely related to the mortality of patients. With the continuous updating of the guidelines, sepsis has been included in the “time window” disease, putting forward a great challenge to the early screening and evaluation of sepsis. This article aims to review the application of Sepsis-Related Organ Failure Assessment, sepsis biomarkers and artificial intelligence algorithms in early screening and evaluation of sepsis, so as to provide guidance tools for timely starting standardized treatment of sepsis.