Monitoring of water vapor in the troposphere is of great importance for real time weather forecasting applications. Over the past decade, Global Navigation Satellite Systems (GNSS) have become an effective supporting tool for deriving reliable and accurate weather forecasts. The main objective of this study is to investigate the accuracy of precipitable water vapor (PWV) estimation by using GNSS observations based on Global Pressure and Temperature empirical (GPT2w) model meteorological parameters (Water vapor weighted mean temperature; T-m, Pressure; P). PWV values derived from GNSS observations based on surface meteorological data were accepted as true values for validation. PWV values were obtained during humid and dry periods for seven days on globally distributed three reference stations. In order to evaluate the regional effect, three reference stations were selected in Australia, Turkey and Greenland. Root mean square error (RMSE) of PWV estimation based on meteorological data derived from GPT2w was found 1.8 mm for overall assessment.