The aim of this study is to investigate the contamination of Cronobacter sakazakii in ready-to-eat foods， and to analysis the epidemic strains isolated in China. Ten strains were isolated from ready-to-eat foods， and the identification was performed by use of MALDI-TOF-MS， VITEK 2 and 16s rDNA sequence analysis. Multilocus sequence typing (MLST) method was employed for genotyping of the isolates. The phylogenetic tree was constructed by analyzing the sequences of the strains which were isolated in China in PubMLST database. MALDI-TOF-MS results revealed that the protein peak of 5740 m/z is the specific protein in C. sakazakii， and there are some differences in the proteins peaks of 5380 m/z， 6255 m/z， 7159 m/z which provide basic data for further genotyping by MALDI-TOF-MS. There were ten C. sakazakii strains identified from 163 isolates. MLST analysis results demonstrated that there were four sequence types (ST21， ST73， ST60， ST7) of the 10 isolates and the main resources of the isolates were cereal foods. The main resources of the isolates in China were clinical， plant food and infant formula. The ST21 strains were more epidemic than the others in China. This study provide basic data for source tracking， epidemiological investigating and control of C. sakazakii， it is useful in reducing the dissemination risk of C. sakazakii in ready-to-eat foods.