Direct measurements of CH4 and CO2 atmosphere/soil exchange with a drained upland inceptisol were made over a 2-year period in a mixed hardwood forest in New Hampshire. Soil gas concentrations of CH4 and CO2 were also monitored over the same period. Soil incubation experiments were used to characterize the depth variation and the temperature response of the consumption and production reactions. CH4 was always taken up by the soils after spring thaw in April. Maximum rates of consumption were 4.8 mg CH4 m(-2) d(-1) in 1989 and 4.9 mg CH4 m(-2) d(-1) in 1990. CO2 efflux was much higher with rates over 25 g CO2 m(-2) d(-1) measured in July 1990. Annually, about 600 mg CH4 were consumed and approximately 2 kg CO2 emitted per square meter. There was an apparent negative correlation between soil CH4 and CO2 concentrations. Gas samples taken in the surface litter layer showed CH4 to be depleted and CO2 enhanced within 1 cm of the surface. Below the surface, CH4 decreased and CO2 increased with depth. At 15 cm, CH4 was never greater than 0.3 ppm and was usually less than 0.2 ppm. At the same depth, CO2 concentrations ranged from 1100 to over 7500 ppm. Incubation experiments indicated that CO2 was being produced throughout the top 15 cm of surface soil with a similar temperature response. Significant CH4 oxidation was measured only in a zone at the top of the mineral soil layer. Both activities could be stopped by autoclaving. CO2 flux from the ground throughout the year was driven by biological activity, mainly soil and root respiration. CH4 uptake, on the other hand, was more complicated. Biological activity controlled the establishment of soil concentration gradients, and so in spring, CH4 influx was tightly linked to rates of consumption. However, in summer and fall, diffusive supply of CH4 to its site of consumption in the soils limited flux rates.