BACKGROUND: Natural microbial communities are extremely complex and dynamic systems in terms of their population structure and functions. However, little is known about the in situ functions of the microbial communities. RESULTS: This study describes the application of proteomic approaches (metaproteomics) to observe expressed protein profiles of natural microbial communities (metaproteomes). The technique was validated using a constructed community and subsequently used to analyze Chesapeake Bay microbial community (0.2 to 3.0 microm) metaproteomes. Chesapeake Bay metaproteomes contained proteins from pI 4-8 with apparent molecular masses between 10-80 kDa. Replicated middle Bay metaproteomes shared approximately 92% of all detected spots, but only shared 30% and 70% of common protein spots with upper and lower Bay metaproteomes. MALDI-TOF analysis of highly expressed proteins produced no significant matches to known proteins. Three Chesapeake Bay proteins were tentatively identified by LC-MS/MS sequencing coupled with MS-BLAST searching. The proteins identified were of marine microbial origin and correlated with abundant Chesapeake Bay microbial lineages, Bacteroides and alpha-proteobacteria. CONCLUSION: Our results represent the first metaproteomic study of aquatic microbial assemblages and demonstrate the potential of metaproteomic approaches to link metagenomic data, taxonomic diversity, functional diversity and biological processes in natural environments.