This study investigated the air- and oxy-combustion characteristics of liquid fuels (diesel oil, kerosene, and heavy oil), using a computational fluid dynamics (CFD) approach. Various key aspects of the combustion characteristics of these liquid fuels in a down-fired laboratory furnace are presented. The flow characteristics, flame structure, fuel evaporation, and the formation of CO in turbulent nonpremixed flames with different O-2/CO2 fractions are discussed in detail. The results of oxy-combustion are also compared with air combustion. Three cases of oxy-combustion (i.e., OF21, OF30, and OF35, with oxygen contents of 21%, 30%, and 35% (by volume), respectively) are considered. Evaporation rates were reduced when N-2 in the air was replaced by CO2 in oxy-combustion; however, similar evaporation rates are obtained when the volume of O-2 in oxy-combustion was increased to 30%. Combustion temperature decreased when N-2 in the air was replaced by CO2 in the oxycombustion environments at the same mole fraction. However, when the O-2/CO2 mole fraction was increased, the temperatures were similar to that of the air-combustion environments. Moreover, because of better evaporation of the fuel and the higher temperatures attained in oxy-combustion, the flame length decreased. By contrast, oxy-combustion yields high CO concentrations compared with the air-combustion environments. The CO concentrations decreased when oxygen content in the oxy-combustion cases was increased. In addition, among the three fuels considered, heavy oil predicted the highest CO concentrations, while diesel and kerosene were in a comparable range. Furthermore, soot concentrations are found to be lower in oxy-combustion, compared to air-combustion environments.