An existing model of C and N dynamics in soils was supplemented with a plant growth submodel and cropping practice routines (fertilization, irrigation, tillage, crop rotation, and manure amendments) to study the biogeochemistry of soil carbon in arable lands. The new model was validated against field results for short-term (1–9 years) decomposition experiments, the seasonal pattern of soil CO2 respiration, and long-term (100 years) soil carbon storage dynamics. A series of sensitivity runs investigated the impact of varying agricultural practices on soil organic carbon (SOC) sequestration. The tests were simulated for corn (maize) plots over a range of soil and climate conditions typical of the United States. The largest carbon sequestration occurred with manure additions; the results were very sensitive to soil texture (more clay led to greater sequestration). Increased N fertilization generally enhanced carbon sequestration, but the results were sensitive to soil texture, initial soil carbon content, and annual precipitation. Reduced tillage also generally (but not always) increased SOC content, though the results were very sensitive to soil texture, initial SOC content, and annual precipitation. A series of long-term simulations investigated the SOC equilibrium for various agricultural practices, soil and climate conditions, and crop rotations. Equilibrium SOC content increased with decreasing temperatures, increasing clay content, enhanced N fertilization, manure amendments, and crops with higher residue yield. Time to equilibrium appears to be one hundred to several hundred years. In all cases, equilibration time was longer for increasing SOC content than for decreasing SOC content. Efforts to enhance carbon sequestration in agricultural soils would do well to focus on those specific areas and agricultural practices with the greatest potential for increasing soil carbon content.