Natural abundance of ¹³C is often used to investigate C turnover in soil after a change of the C₃ vegetation to C₄ crops. Most studies have been carried out with maize as a very common agriculturally used annual C₄ plant. Focussing research only on maize may result in biased information for other cultivated C₄ plants. To study C dynamics in soil, we applied ¹³C natural abundance approach by cultivation of the perennial C₄ energy grass Miscanthus x giganteus. Such investigations are especially important because growing of energy grasses is assumed as a measure to sequester C in soil. Studying dynamics of individual C pools provide more specific information about their turnover rates and C sequestration potential compared to bulk soil analyses. Microbial biomass is one of such C pools, which has fast turnover. Using ¹³C natural abundance approach after C₃ to C₄ vegetation change allows to estimate turnover rates of microbial biomass. Coupling microbial biomass turnover with CO₂ efflux and its C isotopic composition allows evaluation of contribution of new (C₄) and old (C₃) soil organic matter to test the decomposability of SOM pools. Many studies with maize showed, that soil texture strongly affects C turnover. Aim of this study was to test the effect of soil texture on mineralisation of old C (>12 years) and new C (<12 years). We used different textured soils with Miscanthus cultivation for 12 years and 1) determined the contribution of Miscanthus-derived C to microbial biomass and 2) compared the mineralisation of old C₃-C and new C₄-C (Miscanthus-derived C) by incubation under controlled conditions. Samples were taken from 0-10 cm and 10-20 cm depth from the Miscanthus and from neighboured grassland plots from three soils textured as 1) fine loamy sand, 2) coarse loamy sand, or 3) silty loam. δ¹³C values of total SOM and microbial biomass estimated by chloroform-fumigation-extraction were measured to calculate the portion of Miscanthus-derived C. Soil samples were incubated for half a year in air-tied glass vessels at 20 °C and 70% water holding capacity. Microbial biomass and its δ¹³C value were measured 3 weeks after the start and at the end of the incubation. Evolved CO₂ was trapped in 1 M NaOH and changed twice a month during the incubation. Total evolved CO₂ trapped in NaOH was measured by titration and its δ¹³C value was determined after precipitation as SrCO₃ with IRMS. Soil texture affected the incorporation of Miscanthus-derived C into microbial biomass as well as the mineralisation of C₃-C and Miscanthus-derived C₄-C. The content of microbial biomass was lower in sandy soils, but it contained higher portions of Miscanthus-derived C compared with loamy soil. Poor aggregation and lower clay content in sandy soil led to a faster mineralisation of soil organic carbon (SOC) than in loamy soil. Mineralisation rate of total SOC decreased significantly in all soils with increasing incubation time and the differences between the soils were alleviated. In all soils mineralisation of new Miscanthus-derived C was faster than the mineralisation of old C (C₃), which was better stabilized. Our hypothesis regarding the effect of soil texture was confirmed by CO₂ efflux and its C isotopic composition. General statement about the possibility of C sequestration in soil by cultivation of energy grasses cannot be confirmed without further investigations of the individual locations. Incorporation of new C₄-C into microbial biomass was 1.5 to 3 times higher than C₄-C content in total SOM.