Originality StatementThe ecological stoichiometry of microbial biomass has most often focused on the ratio of the biologically-important elements carbon (C), nitrogen (N), and phosphorus (P) and has primarily been examined at a resolution where the contribution of the individual is masked by the reported population or community average. However, reporting population or community averages makes it difficult to assess phenotypic plasticity and stochasticity and mask important information required to understand both the drivers and implications of microbial biomass stoichiometry in nature. One way to assess the diversity of individual microbial phenotypes is through the use of single-cell techniques such as energy dispersive spectroscopy (EDS). EDS reports cellular quotas for the majority of elements composing microbial biomass including C, N, and P. In this study, by measuring individual cells within a microbial community or population, we describe for the first time the stoichiometry of microbial biomass as a distribution instead of an average. Exploration of stoichiometric trait distributions (as presented here) has the potential to improve our understanding of how nutrients interact with individual microorganisms to structure the elemental content of bacterial biomass and better describe how bacterial community biomass affects the ecosystems within which these organisms exist.SummaryTo assess the potential for EDS to describe the stoichiometric variance within populations and communities we measured the stoichiometric trait distribution of cultured freshwater bacterial populations under different resource conditions and compared them to natural microbial communities sampled from three lakes. Mean biomass C:N:P values obtained by EDS matched closely to those obtained by bulk measures using traditional analytical techniques for each freshwater isolate. However, we observed pronounced differences in the stoichiometric trait distributions of freshwater bacterial isolates compared to the stoichiometric trait distributions of natural communities. The stoichiometric trait distribution of the environmental isolates changed with P availability, growth phase, and genotype, with P availability having the strongest effect. The distribution of biomass ratios within each isolate growth experiment were the most constrained during stages of rapid growth and commonly had unimodal distributions. In contrast to the population distributions, the distribution of N:P and C:P for a similar number of cells from each of the mixed lake communities had narrower stoichiometric distributions and more commonly exhibited multiple modes.