A linear programming technique was used to estimate the revenue maximizing allocation of land for a representative household using conservation agriculture production system (CAPS) and farmers’ traditional practices. The model was optimized in five different scenarios. Scenario 1, 2, 3 and 4 were build by allowing annual soil loss to 1, 2, 3 and 4 t ha-1 yr-1 respectively, whereas scenario 5 was build with unconstrained soil loss. Scenario suggested that unless soil loss is considered, conservation tillage does not appear in the profit maximizing allocation of land. Practice with strip tillage appeared in the profit optimized model of all scenarios where soil loss was constrained. Scenario 1 and 2 had about 71 and 66 % of land allotted to maize followed by millet+cowpea intercrop with strip tillage practice. Result also suggested that the representative farm have to sacrifice about $88.6 for about 7 years and $50.1 ha-1 yr-1 (-7.6% and -4.1% revenue) if they aim to reduce the soil loss to 1 and 2 t ha-1 yr-1 respectively. An analysis of the total change in economic surplus associated with adopting the revenue maximizing crop mix was completed. The analysis suggests that conservation agriculture will eventually pay off because total change in economic surplus for 12 years is estimated to be $3,735 million (net present value) if only 1% of the total area adopts the revenue maximization crop mix with a 2 t ha-1 yr-1 soil loss constraint.