Soil CO2 efflux (soil respiration) plays a crucial role in the global carbon cycle and efflux rates may be strongly altered by climate change. We investigated the spatial patterns of soil respiration rates in 144 measurement locations in a 0.5-ha plot and the temporal patterns along a 300-m transect in the 0.5-ha plot. Measurements were made in an unmanaged, highly heterogeneous beech forest during 2000 and 2001. We investigated the effects of soil, roots and forest stand structure on soil respiration, and we also assessed the stability of these spatial patterns over time. Soil temperature alone explained between 68 and 95% of the temporal variation in soil respiration; however, pronounced spatial scatter of respiration rates was not explained by soil temperature. The observed spatial patterns stayed remarkably stable throughout the growing season and over 2 years. The most important structural parameter of the stand was the mean diameter at breast height of trees within a distance of 4 m of the measurement locations (m-dbh4), which explained 10-19% of the variation in soil respiration throughout the growing season. Among the soil chemical parameters, carbon content (bulk as well as dissolved) and magnesium content explained 62% of the spatial variation in soil respiration. The final best model combining soil, root and stand structural parameters (fine root biomass, soil carbon content, m-dbh4 and soil water content) explained 79% of the variation in soil respiration, illustrating the importance of both biotic and abiotic factors.