Aims The theory of ecological stoichiometry mostly builds on studies of natural terrestrial ecosystems, whereas only limited stoichiometry information is available in response to agronomic practices. Methods We designed a greenhouse experiment in order to disentangle the specific role of cover crop identity and soil characteristic in affecting nutrient stoichiometry of a plant-microbe-soil system. Results Nutrient ratios of cover crop biomass were species-specific and the growth rate explained, for most species considered, the stoichiometric differences in response to soil type. In contrast, the nutrient stoichiometry of soil microbes was more homeostatic and did not respond to either cover crop identity or soil type. Compared to bare soil, the presence of cover crop enhanced microbial phosphorus immobilization in the clay-rich soil, whereas it promoted microbial carbon biomass and microbial nitrogen immobilization in the sandyrich soil. A greater microbial cumulative respiration in clay soils, where a higher microbial biomass C at the beginning of the incubation was observed, suggested a major role of soil type, compared to cover crop identity, in affecting microbial metabolism. Conclusions By understanding the stoichiometric constraints in the plant-microbe-soil system, our findings can help to implement agro-ecological practices by selecting appropriate cover crop species in relation to soil type in order, for example, to avoid nutrient limitation due to microbial nutrient immobilization.