Abstract: The study of land use change and the evolution of ecosystem service value (ESV) is of great significance for optimizing the allocation of land resources in fragile karst mountainous areas and realizing ecological restoration. Taking Weng'an County in Guizhou Province as an example, the remote sensing image data from 2001, 2011, and 2019 were used to predict the land use pattern in 2027 through CA-Markov model, and quantitatively analyze the spatio-temporal evolution trend of land use and ESV in the study area from 2001 to 2027. The results showed that the land use types in Weng'an County were mainly forest land and arable land. The transfer of arable land to forest land was relatively intense. The area of forest land and construction land continued to increase, and the dynamic degree of construction land and water body was the largest; ESV was generally increasing, and the transfer of arable land to forest land was the main contributor to the increase of ESV. The increase of ESV mainly occurred in the western and northeastern regions. In the future, low ESV areas would still be distributed in the central urban areas; ESV generated by the service type was mainly based on soil formation and protection services, followed by water sources conservation service and biodiversity protection; forest land and cultivated land were sensitive factors that caused ESV changes in the study area, and ESV of the study area was inelastic to improve the value coefficient. During the study period, the land use structure of Weng'an County had changed significantly, and the ESV increased significantly.