Clostridium difficile has emerged as a major cause of infectious diarrhea in hospitalized patients, with increasing mortality rate and annual healthcare costs exceeding $3 billion. Since C. difficile infections are associated with the use of antibiotics, there is an urgent need to develop treatments that can inactivate the bacterium selectively without affecting commensal microflora. Lytic enzymes from bacteria and bacteriophages show promise as highly selective and effective antimicrobial agents. These enzymes often have a modular structure, consisting of a catalytic domain and a binding domain. In the current work, using consensus catalytic domain and cell-wall binding domain sequences as probes, we analyzed in silico the genome of C. difficile, as well as phages infecting C. difficile. We identified two genes encoding cell lytic enzymes with possible activity against C. difficile. We cloned the genes in a suitable expression vector, expressed and purified the protein products, and tested enzyme activity in vitro. These newly identified enzymes were found to be active against C. difficile cells in a dose-dependent manner. We achieved a more than 4-log reduction in the number of viable bacteria within 5 h of application. Moreover, we found that the enzymes were active against a wide range of C. difficile clinical isolates. We also characterized the biocatalytic mechanism by identifying the specific bonds cleaved by these enzymes within the cell wall peptidoglycan. These results suggest a new approach to combating the growing healthcare problem associated with C. difficile infections.