The innovation of spatial poverty measurement method is the key to precisely allocate poverty alleviation resources and improve the quality of rural poverty alleviation. This paper summarized the three-dimensional analysis framework of spatial poverty and constructed the spatial poverty index system at county scale and the method of geographical identification. With Jingyuan county as a study case, the geographical identification of spatial poverty was carried out. The results can be obtained as follows. (1) Spatial poverty index (SPI) in different geographical areas is listed in the order: Plain Area in Valleys Formed by Erosion and Deposition of Rivers (PAVFEDR) (mean value 1.571) > Denudation Structure in Hilly Regions (DSHR) (mean value -0.199) > Erosion Structure in Mountainous Regions (ESMR) (mean value -0.334), indicating that the highest poverty degree is in ESMR, which should be the main area of poverty alleviation. The common feature of spatial poverty in the three geographical areas is economic disadvantage (C4-C7, C17). The main differences are: for PAVFEDR, ecological disadvantage (C25, C16) > economic disadvantage (C5) > location disadvantage (C20) > political disadvantage (C8); for DSHR, economic disadvantage (C5-C7, C17) > location disadvantage > (C18-C20) > ecological disadvantage (C16, C23-C25) > political disadvantage (C8); for ESMR, economic disadvantage (C4-C7, C17) > political disadvantage (C8) > location disadvantage (C18-C20) > ecological disadvantage (C16, C21-C25). (2) The rank of SPI for different ethnic villages is as follows: Han villages (mean value 1.484) > Hui villages (mean value 1.262) > Hui-Han mixed villages (mean value 1.033), indicating that Hui-Han mixed villages at county scale should be the key villages for poverty alleviation. The common feature of spatial poverty in different ethnic villages is the economic disadvantage resulting from the long distance to the nearest market (C17), as well as the ecological disadvantage due to less arable land per capita (C23) and small crop area, especially economical crop area (C24). The main differences of spatial poverty in different ethnic villages are ecological disadvantage resulting from poor living conditions and high disaster loss rate in Han villages, as well as economic disadvantage caused by lack of loans (C5). For Hui villages, it is economic disadvantage caused mainly by low level of knowledge (C12, C13) that leads to the high opportunity cost in entering the market. In Hui-Han mixed villages the differences of production and lifestyle resulted in difficulties in allocation of poverty alleviation resources, consequently the satisfaction rate of poverty alleviation policies (C8) is the lowest for farmers, reflecting that the political disadvantage of spatial poverty is most significant. (3) There are significant differences in spatial poverty among different ethnic villages in the same physical geographical area. The mean values of SPI for Han, Hui, and Hui-Han mixed villages in PAVFEDR are 0.526, 2.557 and 1.644, respectively, which indicates that the Han villages have a high level of poverty and economic disadvantage (C5); the mean values of SPI for Hui and Hui-Han mixed villages in DSHR are 0.321 and -1.934, respectively, which indicates that the poverty level for Hui-Han mixed villages is high, and economic disadvantage (C2, C6, C7) and ecological disadvantage (C16) are both significant; the mean values of SPI in Han, Hui and Hui-Han mixed villages in ESMR are 1.031, -0.029 and -0.842, respectively, which indicates that the poverty level of Hui-Han mixed villages is high, and economic disadvantage (C5-C7, C17), political disadvantage (C8), location disadvantage (C18-C20) and ecological disadvantage (C16, C25) are all present. Therefore, the Han-villages in PAVFEDR, Hui-Han mixed villages in both DSHR and ESMR should be the key villages of poverty alleviation.