The effect of clear-fell harvesting on soil greenhouse gas (GHG) fluxes of carbon dioxide (CO.sub.2 ), methane (CH.sub.4 ), and nitrous oxide (N.sub.2 O) was assessed in a Sitka spruce forest growing on a peaty gley organo-mineral soil in northern England. Fluxes from the soil and litter layer were measured monthly by the closed chamber method and gas chromatography over 4 years in two mature stands, with one area harvested after the first year. Concurrent measurements of soil temperature and moisture helped to elucidate reasons for the changes in fluxes. In the 3 years after felling, there was a significant increase in the soil temperature, particularly between June and November (3 to 5 .sup." C higher), and in soil moisture, which was 62 % higher in the felled area, and these had pronounced effects on the GHG balance in addition to the removal of the trees and their carbon input to the soil. Annual soil CO.sub.2 effluxes reduced to almost a third in the first year after felling (a drop from 24.0 to 8.9 t CO.sub.2 ha.sup.-1 yr.sup.-1) and half in the second and third year (mean 11.8 t CO.sub.2 ha.sup.-1 yr.sup.-1) compared to before felling, while those from the unfelled area were little changed. Annual effluxes of N.sub.2 O more than doubled in the first two years (from 1.0 to 2.3 and 2.5 t CO.sub.2 e ha.sup.-1 yr.sup.-1, respectively), although by the third year they were only 20 % higher (1.2 t CO.sub.2 e ha.sup.-1 yr.sup.-1). CH.sub.4 fluxes changed from a small net uptake of -0.03 t CO.sub.2 e ha.sup.-1 yr.sup.-1 before felling to a small efflux increasing over the 3 years to 0.34 t CO.sub.2 e ha.sup.-1 yr.sup.-1, presumably because of the wetter soil after felling. Soil CO.sub.2 effluxes dominated the annual net GHG emission when the three gases were compared using their global warming potential (GWP), but N.sub.2 O contributed up to 20 % of this. This study showed fluxes of CO.sub.2, CH.sub.4, and N.sub.2 O responded differently to clear-felling due to the significant changes in soil biotic and abiotic factors and showed large variations between years. This demonstrates the need for multi-year measurements of all GHGs to enable a robust estimate of the effect of the clear-fell phase on the GHG balance of managed forests. This is one of very few multi-year monitoring studies to assess the effect of clear-fell harvesting on soil GHG fluxes.