Abstract: Improving the utilization efficiency of cultivated land is an important initiative to effectively promote the transformation of agricultural modernization in Karst mountainous areas. This study evaluates the cultivated land use efficiency (CLUE) across 88 counties (cities/districts) in Guizhou Province from 2000 to 2020 by the super-efficiency Slack-Based Measure (SBM) model. Through cold/hot spot analysis, trend surface analysis, and geographically and temporally weighted regression (GTWR) modeling, this paper systematically investigated the spatiotemporal evolution patterns and influencing factors. Results show: 1) the provincial CLUE remains suboptimal, with annual efficiency values ranging between 0.5-0.6, failing to reach the effective production frontier. A fluctuating downward trend characterizes the overall efficiency trajectory; 2) distinct spatiotemporal differentiation patterns emerge, where high-efficiency zones demonstrate pronounced spatial clustering characteristics, forming both concentrated clusters and linear belt-shaped aggregations with clearly demarcated hierarchical boundaries; 3) spatial distribution follows an inverted U-shaped spatial gradient progressive trend from west to east and south to north. The north-south differentiation (manifested as“ higher north, lower south” ) proves more significant than the east-west variation (“ higher east, lower west” ), indicating strong spatial orientation; and 4) multivariate analysis identifies five key determinants: average slope gradient, per capita cultivated land area, agricultural mechanization level, per capita GDP, and urbanization rate. These factors exhibit multi-scale interactions and spatiotemporal heterogeneity in their influence mechanisms. The findings offer valuable decision-making references and scientific evidence for optimizing land utilization, enhancing ecological governance, and advancing agricultural modernization in Karst mountainous regions.