In view of the shortcomings of the existing research which has split the basic change process of ecological land loss. Based on the land use data of the Beijing-Tianjin-Hebei urban agglomeration in 1984, 2000 and 2015, the trajectory analysis method was used in this paper to identify the temporal and spatial dynamics of land use change in the study area. In other words, the spatial correlation characteristic of ecological land loss in the Beijing-Tianjin-Hebei urban agglomeration was explored from the perspective of process integrity. On this basis, the spatial autocorrelation analysis method was used to explore the spatial autocorrelation pattern of forest loss, grassland loss and water loss at different scales, and find out the high-incidence area of ecological land loss in this region. At the same time, from the perspective of the integration of the eco-environment of the urban agglomeration, the exploratory suggestions or countermeasures against the loss of ecological land use was proposed. The results showed that (1) From 1984 to 2015, the main changes in land use of the study area were the loss of forest, grassland, water and farmland, and 40% of the lost area turned into artificial surface. (2) The spatial autocorrelation of forest loss, grass loss, and water loss increased with increasing spatial scale. In the 7 km × 7 km spatial scale, There was a strong spatial positive correlation between the three types of ecological land loss, and the spatial positive correlation gradually decreased with the increase of the distance threshold. (3) Under the distance threshold of 10 km, the high incidence areas of forest loss and grassland loss were concentrated in the northwestern part of Beijing-Tianjin-Hebei urban agglomeration, and the high incidence areas of water loss were concentrated near the eastern Bohai Bay. From the perspective of coordinated development of the study area, the eco-environmental protection should resolve differences and unbalanced development and promote the integrated development of this urban agglomeration.