There has been considerable debate on whether root/rhizosphere respiration or bulk soil respiration is more sensitive to long-term temperature changes. We investigated the response of belowground respiration to soil warming by 3 °C above ambient in bare soil plots and plots planted with wheat and maize. Initially, belowground respiration responded more to the soil warming in bare soil plots than in planted plots. However, as the growing season progressed, a greater soil-warming response developed in the planted plots as the contribution of root/rhizosphere respiration to belowground respiration declined. A negative correlation was observed between the contribution of root/rhizosphere respiration to total belowground respiration and the magnitude of the soil-warming response indicating that bulk soil respiration is more temperature sensitive than root/rhizosphere respiration. The dependence of root/rhizosphere respiration on substrate provision from photosynthesis is the most probable explanation for the observed lower temperature sensitivity of root/rhizosphere respiration. At harvest in late September, final crop biomass did not differ between the two soil temperature treatments in either the maize or wheat plots. Postharvest, flux measurements during the winter months indicated that the response of belowground respiration to the soil-warming treatment increased in magnitude (response equated to a Q10 value of 5.7 compared with ∼2.3 during the growing season). However, it appeared that this response was partly caused by a strong indirect effect of soil warming. When measurements were made at a common temperature, belowground respiration remained higher in the warmed subplots suggesting soil warming had maintained a more active microbial community through the winter months. It is proposed that any changes in winter temperatures, resulting from global warming, could alter the sink strength of terrestrial ecosystems considerably.