Staged combustion induces the reduction of NO to generate N-2 through the formation of a fuel-rich zone upstream of the flue gas and utilizes the unburned gas by supplying sufficient air for combustion downstream of the flue gas. Since power generation schedules are very constrained and only specific and limited tests can be planned and executed, the use of numerical simulations is currently more suitable for analyzing these large and complex systems. In this study, computational fluid dynamics (CFD) simulation was performed for an industrial-scale fuel oil combustor to determine the effect of staged combustion on NOx emissions. The fuel oil combustor is a 400-MWe opposite-wall unit located in Ulsan, South Korea, where high-sulfur fuel oil (Bunker-C with 2.5% sulfur content) is used. The combustor has a height of 56 m and a cross-sectional area of 10 x12 m(2). Water wall tubes (evaporator) are located on the wall of the lower part of the combustor and sixteen burners are located at four different axial positions. The system is comprised of two superheaters, two reheaters, and an economizer located in the upper part of the combustor. Staged combustion is realized by changing the equivalence ratio of each burner. Under the initial staged combustion conditions adopted by the Ulsan power plant, the concentration of NOx at the exit of the combustor was calculated to be 362 ppm, which was still high even after selective catalytic reduction treatment. However, when more stringent staged combustion conditions were applied, the predicted concentration of NOx decreased to 309 ppm, which is lower than the mandated NOx concentration at the combustor exit.