Strong gradients of resource availability and environmental stress exist through soil profiles, and they are expected to drive soil microbial community size, diversity, and activity.  However the distribution of microbial populations and activities along soil profiles are mostly uncharacterized and thus their relationships to controlling factors are also weakly understood.  This work was towards characterizing bacterial population and activity distributions along a California grassland soil profile (5 to 100 cm) that experiences two seasons: wet and dry.  TRFLP profiles and six clone libraries were prepared from PCR-amplified genes encoding16S rRNA extracted from different soil depths in June and December.  Bacterial community activity was measured by assaying dehydrogenase; microbial biomass was measured by assaying substrate-induced respiration.  Sequence analyses indicated that Acidobacteria, Actinobacteria, Aphaproteobacteria and Betaproteobacteria dominated throughout the soil profile in both seasons.  However, the proportional representation of Acidobacteria, Actinobacteria and Nitrospirae was relatively higher in the subsurface community as compared to the surface.  The DNA yield, microbial biomass, and dehydrogenase activity (DHA) decreased with soil depth, which is consistent with distributions of organic carbon and nitrogen at the study site.  The highest DNA yield for both the surface and subsurface samples occurred during the middle of the wet season.  The microbial biomass and DHA were constant in the deep soils (50 to 100 cm) across the seasons (June and December), which implied that microbial activity was relatively stable in the subsurface soils.  Overall, the distribution patterns of bacterial phylogenetic groups and activity appear to follow depth-dependent patterns in carbon and energy availability as well as environmental stress.  This work addresses which microbial groups account for the major surface to subsurface community shifts, and the relatedness of these shifts to nutrient concentration gradients.