The spatial variation in soil CO2 flux was determined on tallgrass prairie south of Manhattan, KS between May and September 2006. The objective is to more accurately model CO2 interactions for the prairie ecosystem and gain insight on how carbon fluxes influence climate change in Kansas. Fluxes were measured weekly along two 400-m transects using a hand-held Licor LI-8100 CO2 flux analyzer. The sampling transects surveyed a wide range of landforms (e.g., uplands and lowlands), as well as variations in species composition and soil types. Soil collars were installed every 15 m for a total of 28 sampling points per transect. Digital elevation maps were used to describe each sample site. Peak aboveground biomass was determined along the transects in late July. Corrected data were analyzed for temporal stability using Spearman rank and mean relative difference. Results showed that soil respiration demonstrated temporal stability on transects that traversed significant changes in topography with relatively flat uplands correlating with minimal rank stability. Results will help refine prediction of watershed scale soil CO2 flux for modeling carbon interactions in the prairie ecosystem; thereby providing crucial information pertaining to the effect of CO2 (as a greenhouse gas) on global climate change.