Global warming and changes in rainfall amount and distribution may affect soil respiration as a major carbon flux between the biosphere and the atmosphere. The objectives of this study were to investigate the site to site and interannual variation in soil respiration of six temperate forest sites. Soil respiration was measured using closed chambers over 2 years under mature beech, spruce and pine stands at both Solling and Unterlüß, Germany, which have distinct climates and soils. Cumulative annual CO2 fluxes varied from 4.9 to 5.4 Mg C ha−1 yr−1 at Solling with silty soils and from 4.0 to 5.9 Mg C ha−1 yr−1 at Unterlüß with sandy soils. With one exception soil respiration rates were not significantly different among the six forest sites (site to site variation) and between the years within the same forest site (interannual variation). Only the respiration rate in the spruce stand at Unterlüß was significant lower than the beech stand at Unterlüß in both years. Soil respiration rates of the sandy sites at Unterlüß were limited by soil moisture during the rather dry and warm summer 1999 while soil respiration at the silty Solling site tended to increase. We found a threshold of −80 kPa at 10 cm depth below which soil respiration decreased with increasing drought. Subsequent wetting of sandy soils revealed high CO2 effluxes in the stands at Unterlüß. However, dry periods were infrequent, and our results suggest that temporal variation in soil moisture generally had little effect on annual soil respiration rates. Soil temperature at 5 cm and 10 cm depth explained 83% of the temporal variation in soil respiration using the Arrhenius function. The correlations were weaker using temperature at 0 cm (r2 = 0.63) and 2.5 cm depth (r2 = 0.81). Mean Q10 values for the range from 5 to 15 °C increased asymptotically with soil depth from 1.87 at 0 cm to 3.46 at 10 cm depth, indicating a large uncertainty in the prediction of the temperature dependency of soil respiration. Comparing the fitted Arrhenius curves for same tree species from Solling and Unterlüß revealed higher soil respiration rates for the stands at Solling than in the respective stands at Unterlüß at the same temperature. A significant positive correlation across all sites between predicted soil respiration rates at 10 °C and total phosphorus content and C-to-N ratio of the upper mineral soil indicate a possible effect of nutrients on soil respiration.