We conducted a trenching experiment in a mountain forest in order to assess the contribution of the autotrophic respiration to total soil respiration and evaluate trenching as a technique to achieve it. We hypothesised that the trenching experiment would alter both microbial biomass and microbial community structure and that fine roots (less than 2 mm diameter) would be decomposed within one growing season. Soil C0 2 efflux was measured roughly biweekly over two growing seasons. Root presence and morphology parameters, as well as the soil microbial community were measured prior to trenching, 5 and 15 months after trenching. The trenched plots emitted about 20 and 30% less C0 2 than the control plots in the first and second growing season, respectively. Roots died in trenched plots, but root decay was slow. After 5 and 15 months, fine root biomass was decreased by 9% (not statistically different) and 30%, (statistically different) respectively. When we corrected for the additional trenched-plot C0 2 efflux due to fine root decomposition, the autotrophic soil respiration rose to ~26% of the total soil respiration for the first growing season, and to ~44% for the second growing season. Soil microbial biomass and community structure was not altered by the end of the second growing season. We conclude that trenching can give accurate estimates of the autotrophic and heterotrophic components of soil respiration, if methodological side effects are accounted for, only.