Marine organisms are commonly under threat from various pathogens. Edwardsiella tarda is one of the fish pathogens that can infect both cultured and wild fish species. E. tarda can also infect other vertebrates, including amphibians, reptiles, and mammals. Bacteremia caused by E. tarda can be a fatal disease in humans. Blood acts as a pipeline for the fish immune system. Generating blood transcriptomic resources from fish challenged by E. tarda is crucial for understanding molecular mechanisms underlying blood immune response process. In this study, we performed transcriptome-wide gene expression profiling of Japanese flounder (Paralichthys olivaceus) challenged by 8 and 48 h E. tarda stress. An average of 37 million clean reads per library was obtained, and approximately 85.6% of these reads were successfully mapped to the reference genome. In addition, 808 and 1265 differential expression genes (DEGs) were found at 8 and 48 h post-injection, respectively. Gene Ontology (GO) functional enrichment and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis were conducted to search immune-related DEGs. A protein–protein interaction network was constructed to obtain the interaction relationship of immune genes during pathogens stress. Based on KEGG and protein association networks analysis, 30 hub genes were discovered and validated by quantitative RT-PCR. This study represents the first transcriptome analysis based on protein–protein interaction networks in fish and provides us with valuable gene resources for the research of fish blood immunity, which can significantly assist us to further understand the molecular mechanisms of humans and other vertebrates against E. tarda.