Based on a comprehensive analysis of the functional orientation of each city in the Beijing-Tianjin-Hebei urban agglomeration, an urban sustainable development evaluation index system was constructed with four subsystems, including economic development, social development, scientific and technological innovation, and the ecological environment. The entropy method and a back propagation neural network were used to non-linearly measure and classify the sustainable development ability of Beijing-Tianjin-Hebei urban agglomeration during 2006-2015. The results show that: (1) Beijing and Tianjin are at a high level of sustainable development, and sustainable development decreases with increasing distance from Beijing and Tianjin. Handan and Xingtai are at a low level of sustainable development. (2) Beijing's level of sustainable development exhibits a decreasing trend, while that of other cities is increasing. Large cities are associated with greater pressure on sustainable development. (3) Different cities have their own advantages and disadvantages in the different subsystems. Cities should develop targeted development strategies in combination with the advantages and disadvantages of their subsystems.