Abstract: Exploring spatial-temporal variation and influencing factors of rural poverty has important theoretical and practical significance for formulating a reasonable policy to reduce poverty. Taking the incidence of poverty during 2003-2015 in Guizhou Province, this paper makes a quantitative comprehensive analysis of spatiotemporal variation and influencing factors by using the variable coefficient, spatial statistical model, spatial variogram and Bivariate LISA model. Results showed that: 1) The rural poverty in Guizhou Province is deepening as a whole, and the regional differences are increasing. It is spatially high in the east, south, and west, and low in the middle and north. 2) Rural poverty gravity center was between 107.047 2°-107.174 4°E, 26.668 5°-26.713 2°N and was moving to the southeast. The spatial distribution of rural poverty presented a northeast-southwest pattern. 3) The spatial pattern of rural poverty is affected by both random and structural factors. The spatial differentiation is mainly reflected in the south-north direction, while the northwest-southeast direction is more balanced. 4) There is a significant difference in the rural poverty and influencing factors, and the negative correlation between peasant per capita disposable income, per capita grain production and rural poverty is the most obvious. The spatial difference of the Bivariate LISA clustering map between rural poverty and influencing factors are obvious. Finally, from the aspects of coordinating regional coordinated development, formulating differentiated poverty alleviation measures and taking the road of industrial poverty alleviation, this paper also provides a framework for solving rural poverty in Guizhou Province.