Abstract: In this work land desertification sensitivity in the Northern China agro-pastoral transition zone from 2000 to 2023 is assessed using five categories of data (climate, soil, vegetation, land use, and population density)． A total of three models (an improved MEDALUS model, a Spatial Distance Model, and an ecological sensitivity index model, all three each with their individual weight summed) were used to quantify spatiotemporal evolution of desertification sensitivity． A clear spatial pattern is found, with low sensitivity in the southeast and high sensitivity in the northwest． Temporally, the region experienced a trend of decreasing sensitivity from 2000 to 2020, followed by a sharp rebound between 2021 and 2023, characterized by a dramatic expansion of high and very high sensitivity areas． A comparison of the three models reveals high consistency in macro-scale spatial patterns but notable differences in local sensitivity responses． The MEDALUS model and the ecological sensitivity index model exhibited the strongest correlation (r = 0.811) and spatial agreement (Kappa coefficient = 0.642)． This multi-model evaluation provides robust decision-making support for desertification control planning and ecological space management in the region．