The practical challenge in research in the field of gas turbine combustion mainly centres around a clean emission, a low liner wall temperature, a uniform exit temperature distribution for turbomachinery applications along with a fuel economy of the combustion process. An attempt to meet up the challenge has been made in the present paper in developing a computational model based on stochastic separated flow analysis of a typical diffusion controlled spray combustion of liquid fuel in a gas turbine combustor to study the influences of combustor pressure and inlet swirl on the combustion and emission characteristics within the combustor. A standard k-ε model with wall function treatment for near wall region has been adopted for the solution of conservation equations in gas phase. The initial spray parameters are specified by a suitable PDF size distribution and a given spray cone angle. A radiation model for the gas phase, based on first order moment method, has been adopted in consideration of the gas phase as a grey absorbing-emitting medium. Formation of thermal NO x , as a post combustion reaction process, is determined from Zeldovich mechanism.It has been recognized that an increase in swirl number reduces the NO x emission level and improves the pattern factor at all combustor pressures. However, though at lower pressure an increase in swirl number reduces combustion efficiency, the trend is exactly the reverse at higher pressure. With increase in pressure, pattern factor improves while the combustion efficiency falls and the level of NO x emission increases.