Consumption of atmospheric methane and emission of carbon dioxide by soils were measured on unburned and annually burned tallgrass prairie and on adjacent wheat and sorghum agricultural plots in Kansas. Profiles of CH4 and CO2 concentration with soil depth were also measured. Overall patterns of CH4 consumption by soils varied temporally, with soil depth and land use. Mean CH4 consumption for the 200-day sampling period was -1.02 mg CH4 m-2 d-1 (SE=0.13, n=41) for burned prairie, -0.63 (SE=0.09, n=45) for unburned prairie, -0.85 (SE=0.20, n=36) for wheat, and -0.45 (SE=0.08, n=40) for sorghum. Less than 20% of the variance in CH4 consumption was explained by soil temperature and/or moisture content. Overall patterns of CO2 emission from prairie and agricultural soils varied temporally, but not among land use. Mean CO2 emission for the 200-day sampling period was 15.7 g CO2 m2 d-1 (SE=1.8, n=41) for burned prairie, 14.5 (SE=1.3, n=45) for unburned prairie, 13.9 (SE=2.1, n=36) for wheat, and 10.3 (SE=2.1, n=40) for sorghum. More than 70% of the variance in prairie CO2 emission rate was explained by soil temperature and moisture. Crop management practices influenced the timing of CO2 emission from agricultural plots but not the net annual rate of emission. Methane concentrations generally decreased and CO2 concentrations increased with soil depth, and the magnitude of CH4 and CO2 flux generally increased with increased magnitude of the soil gas concentration gradient. Fertilization of agricultural fields had no measured effect on CH4 or CO2 flux or on soil gas concentrations.