The contamination of groundwater in relation to agricultural activities and urban development is one of the most important issues in the management of this valuable resource. Earth statistics and geographic information system techniques can be a strong tool in the production of spatial data and determine the appropriate management strategies. In this study with comparing of Kriging and Co- Kriging series Earth statistics methods, determine the most appropriate method of providing location changes map the amount of groundwater phosphate and nitrate with emphasis on the for drinking. The study area in this research, Gharah- Soo watershed, located in the West of Golestan province. Accuracy evaluation of the results and determine the most appropriate method of interpolation also is done by using criterion of mutual transvaluation and by using criterions of Root Mean Square Error, General Standard Deviation and Mean Absolute Error. Compare of methods represents a high throughput method of Co- Kriging using auxiliary variable, an estimated amount of nitrate and phosphate. In the next step using multiple linear regression, identify factors affecting on the reduction of water quality. Based on the results of the multiple linear regression modeling, independent variables of elevation, soil, distance from land farming, geology, land use, population density and nitrogen fertilizer consumption at the level of 99 percent have significant impact. Distance from residential areas, underground water level and distance of road level also at the level of 99 percent have significant relationship with the distribution of Nitrate. In the case of the phosphate, independent variables of distance from forest, geology and population density, at the level of 99% and the independent variable the relationship between the cultivation area density and the amount of in phosphate fertilizer consumption in level of 95%, have a significant relationship with the distribution of phosphate in Gharah- Soo watershed. The results of the sensitivity analysis of the model with the use of explaining coefficient also this confirms the content. Map Preparation of water quality parameters spatial variations can be in programmed and decisions useful future managers.The contamination of groundwater in relation to agricultural activities and urban development is one of the most important issues in the management of this valuable resource. Earth statistics and geographic information system techniques can be a strong tool in the production of spatial data and determine the appropriate management strategies. In this study with comparing of Kriging and Co- Kriging series Earth statistics methods, determine the most appropriate method of providing location changes map the amount of groundwater phosphate and nitrate with emphasis on the for drinking. The study area in this research, Gharah- Soo watershed, located in the West of Golestan province. Accuracy evaluation of the results and determine the most appropriate method of interpolation also is done by using criterion of mutual transvaluation and by using criterions of Root Mean Square Error, General Standard Deviation and Mean Absolute Error. Compare of methods represents a high throughput method of Co- Kriging using auxiliary variable, an estimated amount of nitrate and phosphate. In the next step using multiple linear regression, identify factors affecting on the reduction of water quality. Based on the results of the multiple linear regression modeling, independent variables of elevation, soil, distance from land farming, geology, land use, population density and nitrogen fertilizer consumption at the level of 99 percent have significant impact. Distance from residential areas, underground water level and distance of road level also at the level of 99 percent have significant relationship with the distribution of Nitrate. In the case of the phosphate, independent variables of distance from forest, geology and population density, at the level of 99% and the independent variable the relationship between the cultivation area density and the amount of in phosphate fertilizer consumption in level of 95%, have a significant relationship with the distribution of phosphate in Gharah- Soo watershed. The results of the sensitivity analysis of the model with the use of explaining coefficient also this confirms the content. Map Preparation of water quality parameters spatial variations can be in programmed and decisions useful future managers.