The heights determined by GPS are ellipsoidal heights and differ from the orthometric heights used in mapping practices. The ellipsoidal heights determined by GPS can be converted into orthometric heights. The geoid heights must be known in order to accomplish the conversion. The most widely used method for determination of geoid heights is the GPS/Levelling method. In this study, in two separate geodetic networks, the geoid heights of test points were interpolated through five different spatial interpolation methods by using the control points whose geoid heights are determined by the GPS/Levelling method and the accuracy of the interpolation methods are evaluated. Four of the applied interpolation methods were deterministic and these were; Original and Modified Shaperd methods of Inverse Distance Weighted (IDW) methods, Polynomial interpolation methods and Multiquadratic interpolation method. The fifth method was, on the other hand, Ordinary Kriging method, which is a geo-statistical method. The difference values were found by subtracting the Nmeasurement geoid height values found through the GPS/Levelling measurement method from the Nestimated geoid height values that are estimated through the interpolation method. The minimum and maximum errors, the mean absolute errors and root mean square error values of the differences have been calculated for each method. The most accurate interpolation method used in geoid modeling has been selected by comparing these values. As a result of the studies, it has been found that the geoid height determination results using the Ordinary Kriging method were more precise compared to the deterministic methods.