This paper takes Anhui section of the Huaihe River Economic Belt as the research object, and comprehensively uses SBM-Undesirable and DEA-Malmquist models to measure the industrial eco-efficiency from 2009 to 2018. With the help of ArcGIS and GeoDA software, the spatial characteristics of industrial eco-efficiency are analyzed. Based on Tobit panel regression model, the paper makes an empirical analysis on the influencing factors of industrial eco-efficiency, and discusses the mechanism of industrial eco-efficiency pattern evolution. The results indicate that: (1) On the whole, the level of industrial eco-efficiency from 2009 to 2018 is relatively low, and the difference of industrial eco-efficiency among eight cities is obvious. Spatially, the eco-efficiency index shows an increasing spatial pattern from north to south. (2) Most of the eight cities in Anhui section are in the input-output development mode of high benefit, medium consumption, low pollution and low benefit, low consumption and low pollution. (3) Economic development level and research innovation intensity have a positive promoting effect on industrial eco-efficiency. Urbanization level and the proportion of environmental protection investment have a negative inhibiting effect on the improvement of industrial eco-efficiency. The industrial structure and the degree of openness have no significant effect on industrial eco-efficiency.