Reduction of environmental pollutants caused by combustion in power plant systems is one of the main challenges for the researchers. To pinpoint the mechanisms of formation and traort of combustion pollutants, it is necessary to have an accurate prediction of temperature field of the combustion products. For this reason, simulation of turbulent combustion flows has attracted much attention in recent years. An appropriate combustion model is required for simulation of these flows. Flamelet model is the most favorite combustion model, due to inherent separation of the turbulent flow field and the chemical reactions. Moreover, the consideration of unsteady flamelet in modeling complex physical phenomena such as radiation heat transfer and slow chemical processes (of pollutants) leads to better results than the steady flamelet assumption. The purpose of this study is to investigate the application of steady and unsteady flamelet models in the simulation of turbulent diffusion bluff body and jet flames. In order to achieve this task, first a database of temperature and chemical species mass fraction is created in steady state. This flamelet library is obtained from the solutions of simple opposed-jet diffusion flames by assuming unity and non-unity Lewis number in the space of mixture fraction. The FlameMaster code is used to build these libraries. Effect of turbulence on the combustion field of flamelets is applied by considering a presumed-pdf (beta-function), which depends only on the mean mixture fraction and its variance. Another input to the library is the mean value of the scalar dissipation rate. By performing the integration of the chemical library over the permissible range of the quantities, the mean values of chemical species mass fraction and temperature is achieved. The transient effects are considered in a post-processing manner using the unsteady flamelet Model. In this model, the transient history of scalar dissipation rate, conditioned at stoichiometric mixture fraction is required. This value is obtained by considering a tracer particle to generate unsteady flamelets. In addition, the probability of finding unsteady flamelet is obtained by solving an unsteady scalar traort equation on a converged solution in the whole domain. Predictions of velocity, mean mixture fraction and its rms using the modified k-? have shown very good agreement with experiment data. Comparisons of the results of steady and unsteady calculations have shown that transient effects do not have much influence on major species, including OH and therefore the structure of the flame can be successfully predicted by steady or unsteady approaches. Hence the predictions of temperature and mass fraction of major species using steady flamelet model are in good agreement with the experimental results. However, NO mass fraction in steady-state simulations using two different chemical mechanisms GRI3.0 and GRI2.11 is over predicted. Furthermore the GRI3.0 mechanism showed overpredictions of NO with both unsteady and steady models. While NO mass fraction in the unsteady flamelet modeling using mechanism GRI2.11and considering radiation have shown good agreement with the experimental data. Thus, unsteady effects are important in slow processes such as the formation of NO and complex phenomena such as radiation heat transfer. Key Words: Turbulent Combustion, Flamelet Modeling, Non-Premixed Flame, Flamelet Library.