In this proposed session, we discuss problems associated with studies of microbial functions and activities in nature, and explore how to integrate different approaches, molecular and process-based measurements. One type of problem centers around working with DNA and RNA from environmental samples. Chris Schadt talks about functional gene microarrays and other techniques for studying functional capability of microbial communities at the level of DNA and mRNA. The arrays allow us to detect the presence and/or expression of genes required for a variety of functional processes including nitrogen, carbon, and sulfur cycling. Matt Wallenstein then discusses application of proteomics to understand microbial physiological responses to stress. Molecular methods can also be useful for studying individual cells in nature. Yuko Hasegawa will talk about how fluorescent in situ hybridization (FISH) combined with spectral imaging analysis can be applied to identify specific groups of microbial cells in the environment. She will also present an application of spectral imaging technology to simultaneously visualize a large number of distinct microbial groups. There are also many problems with measuring the microbial processes. For example, enzyme activity may be out of synchrony with enzyme production and release from microbes. Kurt Smemo uses his work on forest soils to illustrate how production and activity of exoenzymes can vary temporarily, as well as how spatial patterns of enzyme activities and microbial communities affect our ability to link communities and processes. Jeremy Rich critiques approaches previously used to link microbial community structure and ecosystem functioning and points to potentially fruitful avenues for future research. The session ends with two papers on modeling. First, Michael Weintraub discusses modeling approaches to link microbial community composition to ecological processes. Weintraub points out that traditional decomposition models have included neither microbes nor their enzymes as explicit drivers of decomposition. However, when such information is incorporated into models, the behaviors of the models as well as the conclusions that can be drawn from them alter. Finally, Steve Allison talks about integration of microbial communities into large-scale ecosystem models. Allison asks how models can distill the complexity of microbial communities into a manageable set of relevant parameters. Allison also presents a framework for aggregating microbial communities, based on taxonomic and functional genes, and provides examples of how explicit representation of microbial community structure can improve process models.