The invention discloses a method for annotating cell identities based on single cell transcriptome clustering results. The method comprises the following steps: S1, providing survival cells, establishing a library by using 10X genetics, and sequencing to obtain transcriptome sequencing data; S2, filtering the sequencing data obtained in the step S1, then carrying out initial analysis by using thesoftware cellranger, and outputting original data; S3, analyzing the original data output in the step S2; S4, performing cell identity annotation, including the steps: S41, classifying the tag genes of the Cell Marker database according to cell types; S42, sorting the tag genes screened out by the FindAllMarkers function according to a P value; S43, taking an intersection of the tag gene of each type of cells in the Cell Marker database and a cell subset tag gene screened by the FindAllMarkers function, and scoring according to the intersection gene; and S44, sorting according to the score ofthe gene intersection, and annotating the cell type corresponding to the highest score as the cell identity of the current subgroup. By adopting the method for annotating cell identities based on single cell transcriptome clustering results disclosed by the invention, the cells can be accurately and quickly classified and subjected to identity annotation.