Abstract: The dynamic change of tourism eco-efficiency of urban agglomeration in the middle reaches of Yangtze River from 2011 to 2020 was measured and analyzed by combining DEA window analysis method with super-efficiency SBM model. The modified gravity model and social network analysis method were used to explore the evolution of spatial network structure of tourism eco-efficiency in urban agglomeration, and the panel data regression model was constructed to analyze its network effects. The results showed that: (1) The overall mean value of tourism eco-efficiency was 0.719, which showed a downward trend in fluctuation with significant regional differences. (2) The fluctuation of the overall network density and the number of network relations increased to 0.162 7 and 123, and the network level and network efficiency decreased slightly but basically stabilized at 0.2 and 0.8. It that the correlation network structure of tourism eco-efficiency in urban agglomeration  loose, with certain hierarchical characteristics, and the overflow channel was not smooth, which needed to be further optimized. (3) The individual indexes of Wuhan, Changsha, Nanchang, Xiaogan and Yichun were always higher than the mean value, which occupcentral position in the correlation network of tourism eco-efficiency. The individual indicators of Changde, Yiyang, Huanggang, Yichang, Xiangyang and other cities were always lower than the mean value and relatively lag behind, at the edge of the network. (4) The increase of overall network density, the decrease of network level and network efficiency could significantly improve the overall tourism eco-efficiency of urban agglomerations and narrow the regional differences. The improvement of individual network centrality index had a significant promoting effect on urban tourism eco-efficiency.