This thesis evaluates the applicability of forecasting production from low permeability and layered tight gas wells using the Arps hyperbolic equation at earlier times and then switching to the exponential form of the equation at a predetermined minimum decline rate. This methodology is called the minimum terminal decline rate method. Two separate completion types have been analyzed. The first is horizontal completions with multi-stage hydraulic fractures while the second is vertical fractured wells in layered formations, completed with hydraulic fractures. For both completion types both simulated data and real world well performance histories have been evaluated using differing minimum terminal decline rates and the benefit of increasing portions of production history to make predictions. The application of the minimum terminal decline rate method to the simulated data in this study (3 percent minimum decline applied to multiple fractured horizontal wells MFHW- and 7 percent applied to vertical fractured layered wells) gave high errors for some simulations within the first two years. Once additional production data is considered in making predictions, the errors in estimated ultimate recovery and in remaining reserves is significantly reduced. This result provides a note of caution, when using the minimum decline rate method for forecasting using small quantities of production history. The evaluation of real world data using the minimum terminal decline rate method introduces other inaccuracies such as poor data quality, low data frequency, operational changes which affect the production profile and workovers / re-stimulations which require a restart of production forecasting process. Real well data for MFHW comes from the Barnett Shale completions of the type which have been widely utilized since 2004. There is insufficient production history from real wells to determine an appropriate minimum terminal decline rate. In the absence of suitable analogs for the determination of the minimum terminal decline rate it would be impossible to correctly apply this methodology. Real well data for vertical fractured layered wells from the Carthage Cotton Valley field indicate that for wells similar to Conoco operated Panola County wells a feasible decline rate is between 5 percent and 10 percent. Further if a consistent production trend and with more than 2 years of production history are used to forecast, the EUR can be predicted to within plus/minus 10 percent and remaining reserves to within plus/minus 15 percent.