Sepsis is a life-threatening condition in which an uncontrolled inflammatory host response is triggered. The exact pathogenesis of sepsis remains unclear. The aim of the present study was to identify key genes that may aid in the diagnosis and treatment of sepsis. mRNA expression data from blood samples taken from patients with sepsis and healthy individuals was downloaded from the Gene Expression Omnibus database and differentially expressed genes (DEGs) between the two groups were identified. Gene Ontology and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis, and protein-protein interaction (PPI) network construction, was performed to investigate the function of the identified DEGs. Furthermore, for validation of these results, the expression levels of several DEGs were analyzed by reverse transcription quantitative-polymerase chain reaction (RT-qPCR) in three patients with sepsis and three healthy blood samples to support the results obtained from the bioinformatics analysis. Receiver operating characteristic analyses were also used to analyze the diagnostic ability of the identified DEGs for sepsis. The results demonstrated that a total of 4,402 DEGs, including 1,960 upregulated and 2,442 downregulated genes, were identified between patients with sepsis and healthy individuals. KEGG pathway analysis revealed that 39 DEGs were significantly enriched in toll-like receptor signaling pathways. The top 20 upregulated and downregulated DEGs were used to construct the PPI network. Hub genes with high degrees, including interleukin 1 receptor-associated kinase 3 (IRAK3), S100 calcium-binding protein (S100)A8, angiotensin II receptor-associated protein (AGTRAP) and S100A9, were demonstrated to be associated sepsis. Furthermore, RT-qPCR results demonstrated that IRAK3, adrenomedullin (ADM), arachidonate 5-lipoxygenase (ALOX5), matrix metallopeptidase 9 (MMP9) and S100A8 were significantly upregulated, while ectonucleoside triphosphate diphosphohydrolase 1 (ENTPD1) was upregulated but not significantly, in blood samples from patients with sepsis compared with healthy individuals, which was consistent with bioinformatics analysis results. Therefore, AGTRAP, IRAK3, ADM, ALOX5, MMP9, S100A8 and ENTPD1 were identified to have potential diagnostic value in sepsis. In conclusion, dysregulated levels of the AGTRAP, IRAK3, ADM, ALOX5, MMP9, S100A8 and ENTPD1 genes may be involved in sepsis pathophysiology and may be utilized as potential diagnostic biomarkers or therapeutic targets.