Microbial responses to three years of CO2 enrichment (600 μL L–1) in the field were investigated in calcareous grassland. Microbial biomass carbon (C) and soil organic C and nitrogen (N) were not significantly influenced by elevated CO2. Microbial C:N ratios significantly decreased under elevated CO2 (– 15%, P = 0.01) and microbial N increased by + 18% (P = 0.04). Soil basal respiration was significantly increased on one out of 7 sampling dates (+ 14%, P = 0.03; December of the third year of treatment), whereas the metabolic quotient for CO2 (qCO2 = basal respiration/microbial C) did not exhibit any significant differences between CO2 treatments. Also no responses of microbial activity and biomass were found in a complementary greenhouse study where intact grassland turfs taken from the field site were factorially treated with elevated CO2 and phosphorus (P) fertilizer (1 g P m–2 y–1). Previously reported C balance calculations showed that in the ecosystem investigated growing season soil C inputs were strongly enhanced under elevated CO2. It is hypothesized that the absence of microbial responses to these enhanced soil C fluxes originated from mineral nutrient limitations of microbial processes. Laboratory incubations showed that short-term microbial growth (one week) was strongly limited by N availability, whereas P was not limiting in this soil. The absence of large effects of elevated CO2 on microbial activity or biomass in such nutrient-poor natural ecosystems is in marked contrast to previously published large and short-term microbial responses to CO2 enrichment which were found in fertilized or disturbed systems. It is speculated that the absence of such responses in undisturbed natural ecosystems in which mineral nutrient cycles have equilibrated over longer periods of time is caused by mineral nutrient limitations which are ineffective in disturbed or fertilized systems and that therefore microbial responses to elevated CO2 must be studied in natural, undisturbed systems.