The analysis of plant proteomes has drastically expanded in the last few years. Mass spectrometry technology, stains, software and progress in bioinformatics have made identification of proteins relatively easy. The assignment of proteins to particular organelles and the development of better algorithms to predict sub-cellular localization are examples of how proteomic studies are contributing to plant biology. Protein phosphorylation and degradation are also known to occur during plant defense signaling cascades. Despite the great potential to give contributions to the study of plant-pathogen interactions, only recently has the proteomic approach begun to be applied to this field. Biological variation and complexity in a situation involving two organisms in intimate contact are intrinsic challenges in this area, however, for proteomics studies yet, there is no substitute for in planta studies with pathogens, and ways to address these problems are discussed. Protein identification depends not only on mass spectrometry, but also on the existence of complete genome sequence databases for comparison. Although the number of completely sequenced genomes is constantly growing, only four plants have their genomes completely sequenced. Additionally, there are already a number of pathosystems where both partners in the interaction have genomes fully sequenced and where functional genomics tools are available. It is thus to be expected that great progress in understanding the biology of these pathosystems will be made over the next few years. Cheaper sequencing technologies should make protein identification in non-model species easier and the bottleneck in proteomic research should shift from unambiguous protein identification to determination of protein function.