Soil carbon dioxide (CO₂) efflux is one of the largest fluxes within the global carbon cycle. Uncertainty associated with CO₂ flux estimates is of critical importance especially in the compilation of carbon emission inventories. We analyzed factors associated with the spatial and temporal variability of CO₂ efflux in a Piedmont soil. This study was conducted at the North Carolina Agricultural and Technical State University Farm, Guilford County, North Carolina on a Mecklenburg sandy loam (fine, mixed, thermic Ultic Hapludalfs) soil. No-till sweet corn experiment in 2005 summer consisted of two main-plots with cover crop residue and no cover residue, and four sub-plots with varying inorganic nitrogen fertilizer application rates. Specialty melon experiment in 2006 summer was drip irrigated and consisted of the same experimental treatments. The cover crop established in fall was a biculture of cereal rye (Secale cereale L.) and crimson clover (Trifolium incarnatum L.) and was rolled and crimped in the plots prior to sweet corn planting, and was disked and tilled prior to specialty melon planting. Gas samples were collected from the chambers in the field plots that received 62.5 and 125 kg N ha^-1 under cover and no cover residue-amended soils and were analyzed for CO₂ using a portable LI-840 CO₂/H₂O analyzer. A forest site with a pine stand was also included in the study for comparative purposes. Results indicated that the variability of CO₂ flux was significantly higher in cover crop residue-amended plots as compared to no cover plots. In cover plots, CO₂ flux varied greatly with 63 kg N ha^-1 treatment, especially in fall. The CO₂ flux tended to be more variable in no-till sweet corn plots as compared to melon plots with conventional tillage. Influence of micrometeorological parameters on the variability of CO₂ flux and other results will be presented and discussed.