Clostridium perfringens is a well-known food-borne pathogen causing food-borne illnesses. To control this food-borne pathogen in foods, bacteriophage approach could be a good alternative biocontrol agent. US FDA approved phage applications to foods as natural food preservatives since it has been generally recognized as safe (GRAS) in 2006. However, because the consumers generally do not prefer phage additions to foods for the purpose of extension of food preservation, phage-derived endolysin has been studied. Endolysin is a peptidoglycan hydrolase from phage, which breaks down bacterial peptidoglycan. Therefore, it could be used to control specific pathogen by host lysis. It has two domain enzymatically active domain (EAD) and cell wall binding domain (CBD). Fusing CBD and enhanced green fluorescent protein (EGFP), this EGFP-CBD-fused protein could be used for rapid detection. These new biocontrol and rapid detection approaches have attracted the public interest to enhance food safety from contamination of various food-borne pathogens. To obtain a novel endolysin and its CBD, phage CPAS-15 was obtained from American Type Culture Collection (ATCC). The LysCPAS15, an endolysin from phage CPAS-15, was cloned into pET21a vector and expressed in E. coli BL21(DE3). Turbidity reduction assay showed that the purified LysCPAS15 efficiently lysed the host strains of C. perfringens and even C. difficile. In addition, LysCPAS15 is highly stable under various pH (4-8) and temperature conditions (-20-50°C). Furthermore, purified CBD tagged with EGFP (EGFP-CBD-CPAS15) was obtained by cloning and expression using the same E. coli expression system. Mixing and short incubation of this fused protein with the host strain (C. perfringens or C. difficile) showed fluorescence signals on the cell surface, substantiating the host-specific binding activity for rapid detection. In conclusion, this EGFP-CBD-fused protein can be used for rapid detection of these food-borne pathogens, suggesting that CBD can substitute high cost antibody for rapid detection.