Caenorhabditis elegans can be infected by a plethora of pathogens, most of them are also pathogenic for humans. Consequently, the nematode has emerged as a powerful surrogate host to model microbial human infectious diseases in a non-vertebrate, for the study of innate immunity and host-pathogen interactions. Signaling cascades are well investigated that face bacterial or fungal pathogens. We analyzed the downstream processes of these cascades, i.e. the differential expression of effector and regulatory molecules due to a microbial challenge with a pathogenic strain of the bacterium Bacillus thuringiensis (Bt) in comparison to a non-pathogenic Bt strain. The protein abundance profile of the nematode was studied by quantitative proteomics using iTRAQ labeling and 2D-LC-MS analysis. We developed (i) a novel method for the preparation of defined C. elegans samples; (ii) a pooling strategy for fractions in 2D-LC separation schemes; and (iii) an isobaric labeling scheme reducing the number of necessary LC-MS experiments. More than 3,600 proteins were quantified, 288 of which showed altered abundances, implicating protein classes such as lectins, lysozymes, and transthyretin-like proteins to be involved in the nematode innate immune defense. A number of gene products previously only identified by transcriptomic profiling could be verified at the protein level. Moreover, several other protein classes such as proteases, proteins related to autophagy and apoptosis, structural proteins, and proteins involved in chromatin organization were detected. The results provide an overview of the physiological response towards a pathogen at protein level in the important model organism C. elegans, giving insights into highly complex host-pathogen interactions.