dioxide emission by these forest soils. In general, the jack pine sites were stronger methane sinks and had lower carbon dioxide emissions than the black spruce sites. After a few years of recovery, the burned sites tended to be slightly stronger methane sinks than unburned controls. Our results suggest that soil CO2 effluxes from upland black spruce stands may not be immediately impacted by fire, possibly maintained at preburn levels by microbial decomposition of labile compounds released as a result of the fire. By 2 years postfire there appears to be a significant reduction in soil CO2 flux, due to the loss of tree root and moss respiration and possibly to the depletion of fire-related labile compounds. The observed recovery of soil respiration rates to preburn levels by 7 years postburn is probably due to the respiration of regrowing vegetation and the combined effects of elevated soil temperatures (about 4 ø to 5øC warmer than unburned sites) and improved litter quality on soil microbial activities. We estimate that soil CO2 emissions from recently burned boreal forest soils in the northern hemisphere could be of the order of 0.35 Pg C yr -•, which is in good agreement with a previous estimate that was derived in a different manner.