Because of the growing importance of unconventional gas (shale gas, coalbed methane, etc.), gas flow simulation has been receiving greater interest in the industry. Although the real gas pseudo-pressure approach has been available for some time, many commercial simulation tools still use gas pressure as the basis for gas flow calculations. There is a widely held perception that, while the approach can result in faster simulation runs, the two approaches are essentially equivalent or interchangeable. Herein, we show that well productivity is often significantly over-estimated when the standard pressure based approach is used for tight gas reservoirs. The effect is more pronounced at lower pressures.We demonstrate that the difference in results arises from the numerical dispersion in the simulator associated with the pressure formulation. We discuss the implementation within a simulator that models both sub-surface flow as well as flow in a surface network. This allows comparison of simulation runs using the two different formulations.Finally, we present several examples simulating gas flow in tight reservoirs with wells that are horizontally and vertically fractured, as well as a multi-well example. The approach is shown to yield accurate results in contrast to the pressure approach. This has important implications for simulations on a field wide scale and when other phases are present along with gas.This paper shows that the real gas pseudo-pressure approach should be used to simulate tight gas reservoirs, through which the well productivity can be accurately estimated. The implementation approach is also applicable to other commercial simulators in the industry.