A pathogen detection methodology based on Bayesian decision theory has been developed for rapid and reliable detection of Salmonella typhimurium. The methodology exploits principles from statistical signal processing along with impedance spectroscopy in order to analytically determine the existence of pathogens in the target solution. The proposed technique is validated using a cost-effective and portable immunosensor. This device uses label-free, electrochemical impedance spectroscopy for pathogen detection and has been demonstrated to reliably detect pre-infectious levels of pathogen in sample solutions. The detection process does not entail any pathogen enrichment procedures. The results using the proposed technique indicate a detection time of approximately 6 min (5 min for data acquisition, 1 min for analysis) for pathogen concentrations in the order of 500 CFU/ml. The detection methodology presented here has demonstrated high accuracy and can be generalized for the detection of other pathogens with healthcare, food, and environmental implications. Furthermore, the technique has a low computational complexity and uses a minimal data-set (only 30 data-samples) for data analysis. Hence, it is ideal for use in hand-held pathogen detectors. (C) 2008 Published by Elsevier B.V.