Abstract: Spatialization of agricultural output value is a form of spatial statistics critical for analyzing agricultural output value. As a key agricultural province in China, there is the necessity to analyze the factors affecting the spatial distribution characteristics and output value of agriculture in Xinjiang Uygur Autonomous Region. This is critical for adjusting agricultural structure towards high productivity. To analyze agricultural output value on geo-administrative distribution, this paper established a multi-variate regression model with county-based areas of land use types and river lengths as independent variables in combination with area weighted method to spatialize agricultural output value in Xinjiang at 1.0 km spatial resolution. The results of analysis of agricultural output value of Xinjiang in 2012 showed that (1) agricultural output value was closely related to areas of total cultivated land, forest land, grassland and length of local rivers. (2) The suitability of multi-variate regression simulation of agricultural output value was satisfactory with the linearity between the predicted and actual values up to 0.82. (3) The results of the spatial simulation suggested that the overall level of agricultural output value was low in Xinjiang, which was less than 5 000 Yuan per km². High value areas were concentrated in the Ili Region, Tuscaloosa Area, the north slope of Tianshan Mountain, and the upper and middle reaches of Tarim River. The areas with low values were mainly in the south of Taklimakan Desert. The whole distribution of agricultural output value was in zonal or sheet-like pattern. Most of the high value areas were belonged to the sheet-like pattern. The feature of zonal distribution was closely associated with local rivers and roads, but more especially with local river. Zonal distribution areas were mainly along the Tarim River Basin, Kongquehe River, Yeerqiang River Valley and Ili River Valley. The paper further analyzed the relationship between the main roads and agricultural output value. It was noted that the nearer to the roads, the higher were the agricultural output values. (4) The relationship between rivers and agricultural output values illustrated that water resources were the restricting factors of perfect agricultural development in arid areas. (5) The factors affecting the simulation accuracy included the inherent objective conditions of Xinjiang, human factors, regional differences and internal differences in production capacity of land use. Compared with previous studies, this paper considered on-the-ground conditions in arid areas and highlighted the importance of water resources in agricultural development and productivity. This had a real practical value for structural adjustment and optimization of agricultural industry in Xinjiang.