Precipitation is one of the most important climatic factors with high variations in time and space. Determination of the amount of precipitation in different locations is particularly very important. For this reason, to estimate precipitation in various regions of Golestan province, different interpolation methods have been used. Precipitation data of 32 rain gage stations with 26-years period were selected. First the accuracy and homogeneity of data were evaluated by using statistical tests. Then, to determine the best model of spatial pattern of annual precipitation in Golestan province, six methods of IDW, GPI, LPI, RBF, Kriging and Co-Kriging have been used. The criteria of statistical error were used to evaluate the results and to select the most suitable method of interpolation. The results showed that Geostatistics methods were better than deterministic methods, and among the geostatistical models, Kriging give better results (MAE=60.49) than Cokriging (MAE=64.46). Also in cases of using the deterministic methods, RBF and IDW have more accurate results than LPI and GPI. The lowest MAE (60.49) was recorded in Kriging method with spherical model and the highest MAE (288.96) was obtained in GPI method with power 5.