The growth rate hypothesis (GRH) suggests that growth and elemental stoichiometry of organisms are coupled through variation in nucleic acid composition, specifically variation in ribosomal RNA. We examined these interactions in a bacterial community (<1 µm water) from the moderately productive Lake Owasso (Minnesota, USA). A mixed bacterial community was grown in chemostats with different supply carbon:phosphorus ratios (93, 373 and 933 by atoms) and dilution (= growth) rates (0.25, 0.5 and 0.7 h -1 ). We measured bacterial C, N, P, DNA and RNA content to determine resource-and growth-dependent variations in the elemental stoichiometry of bacterial biomass at the community level. In chemostats with high supply C:P ratios (≥ 373), biomass P increased proportionally to growth rate as a result of increased RNA content. High growth rates generated low biomass C:P and N:P ratios. At a low supply C:P (93:1), biomass C:P and N:P increased with increasing growth rate, even though RNA content increased, suggesting that non-nucleic acid P (presumably polyphosphate) was the dominant P-pool in this community and may have been more important in altering biomass stoichiometry at increasing growth rates. Therefore, the coupling of stoichiometry to growth predicted by the growth rate hypothesis was applicable only under P-limited conditions. These chemostat results revealed more complicated responses of bacterial C:N:P stoichiometry at a community level than in individual strains. The Lake Owasso bacterial community was more homeostatic in terms of biomass C:P and N:P stoichiometry than other bacterial communities described in the literature, and the degree of homeostasis for the Lake Owasso bacterial community was similar to that of single bacterial strains. However, few similar measurements have been made, and the degree of biomass C:N:P homeostasis may vary between systems and even within a system during different seasons.