|In order to minimize the danger to the public from pathogen outbreaks, it is imperative to rapidly identify the infectious agent so appropriate response can be mounted and prevent dissemination of the organism. An ideal detection technology should be fast (minutes rather than hours), accurate, have a long shelf life, inexpensive, and relatively easy to use without extensive training or elaborate equipment. The magnetoelastic (ME) biosensors using phage-displayed oligopeptides as probes have the advantage over other technologies because they are fast, cost effective and require minimal sample manipulations. In the continuing effort to advance the phage-ME biosensors, I made several improvements for this technology with focus on phage displayed molecular probe isolation and attachment to the ME sensor platform. Firstly, I implemented a stringent negative selection procedure that enhanced isolation of molecular probes with higher selectivity while minimizing cross reactivity. Secondly, I improved the classic ELISA procedure against live cells to more accurately assess the interaction between molecular probes and bacterial cells. These results demonstrated that liquid phase ELISA yields results with less background and more clearly distinguishes selective probes from less selective probes. Using these two improved approaches, I isolated highly selective phage-displayed oligopeptide probes for detection of an important foodborne pathogen, Salmonella enterica serovar Typhimurium, that recognize the bacterium with 600% higher affinity than other S. enterica serovars, Escherichia coli or Shigella species. In addition, I devised a selection strategy and isolated phage displayed molecular probes that can recognize multiple serovars of pathogenic S. enterica. Based on recent Salmonella data, it is obvious that many S. enterica serovars can cause outbreaks. Thus, these isolated general Salmonella recognition probes can serve as the initial detection technology to demonstrate the presence of many of the common pathogenic serovars of S. enterica including S. Heidelberg, S. Newport, S. Javiana, S. Dublin, S. Braenderup, and S. Montevideo. Furthermore, I developed a genetic scheme to improve immobilization of the phage displayed molecular probes on ME biosensor platforms. The strategy is based on tagging 100% of the phage particles with Strep-tag II and allowing them to bind to streptavidin coated ME platform as a monolayer. The strategy can be used to tag any phage probes with any affinity tag to improve binding of the probes to the sensor platform. Finally, I utilized these improvements to construct a S. Typhimurium specific ME biosensor. I tagged the highly selective S. Typhimurium phage-displayed oligopeptide probe TA1 with Strep-tag II and immobilized it on a 0.028 mm x 0.4 mm x 2 mm magnetoelastic (ME) particle. TA1 biosensor was highly sensitive and I were able to detect between 102–106 bacterial cells. I expect the sensitivity to increase when the TA1 phage probe is used with a smaller ME particle. The TA1 biosensor was highly selective and only recognized S. Typhimurium. When directly compared to the previously used E2 phage biosensor, the TA1 biosensor proved to be more sensitive by 2–4 fold for detection of S. Typhimurium between 102–106 bacterial cells and validated all of the hypotheses and efforts. In summary, the data presented in this dissertation improved the efficacy of a promising ME biosensor technology for accurate and rapid detection of bacterial pathogens.