The experimental study described in this paper is to investigate the control of thermal nitrogen oxides emissions from a 2.28 MW gas-fired test furnace. Tests, including changing axial or radial air flow rate, adding cooling water, and adding staged air, were performed to characterize and op(imize the fuel-rich burning zone and the fuel-lean burnout zone independently. Detailed measurements of O_2, CO_2, CO, NO and NOx were made at the fuel-rich burning zone and furnace exit. The influence of forming CO, NO and NOx was examined. Results indicated that adding staged air in the fuel-rich burning zone (75 cm from burner) will reduce the mawximum NO and NO_x emissions. Adding cooling water in a right position may further lower the NO and NO_x emissions. In addition, the least formation of thermal nitrogen oxides in the first stooge fuel-rich bunting zone will occur at the stoichiometric ratio’s inverse value, (φ_1)~(-1), 0.65 to 0.7. The experimental study described in this paper is to investigate the control of thermal nitrogen oxides emissions from a 2.28 MW gas-fired test furnace. Tests, including changing axial or radial air flow rate, adding cooling water, and adding staged air, were performed to Characterize and op (imize the fuel-rich burning zone and the fuel-lean burnout zone independently. Detailed measurements of O_2, CO_2, CO, NO and NO x were made at the fuel-rich burning zone and furnace exit. The influence of forming CO , NO and NO x was examined. Results indicated that adding staged air in the fuel-rich burning zone (75 cm from burner) will reduce the mawximum NO and NO_x emissions. Adding cooling water in a right position may further lower the NO and NO_x emissions . In addition, the least formation of thermal nitrogen oxides in the first stooge fuel-rich bunting zone will occur at the stoichiometric ratio’s inverse value, (φ_1) -1, 0.65 to 0.7.