Grassland plays an important role in controlling rocky desertification and supporting animal husbandry in karst areas. Information about grassland spatial distribution patterns and driving forces of change in grassland patterns in karst areas is important for planning maintenance of regional grassland ecological functions and achieving sustainable development. Based on land-use datasets, in this study, we analyzed the spatial patterns and changes of grassland distribution in Guizhou Province from 1980 to 2020. Combined the landscape pattern and spatial autocorrelation to deeply identified the evolution law of grassland spatial distribution patterns and effective management areas. And quantified the driving forces of the spatial distribution patterns evolution of grassland by using geographic detector. It was found that: 1)In the past 40 years, the change of grassland area in Guizhou Province can be divided into three stages: a growth period (1980-2000), a decline period (2000-2015) and a recovery period (2015-2020), with an overall decrease from 1980 to 2020 of 176.88 km2. The areas that have undergone changes mainly occurred in the western and southern regions of Guizhou Province, with the main changes being transfer between grassland, forest, and cultivated land. The overall spatial occurrence of grassland could be summarized as “high in the west and south, low in the east and north”. 2)Over time from 1980 to 2020, the degree of fragmentation of grassland increased, the degree of aggregation decreased, and the shape tended to become complex. Grassland patches at the county level are more fragmented and scattered, but their shapes are more regular, than at provincial level. 3)The Anselin Local Moran’s I tool was used to assess the uniformity of aggregation. Using this analytical approach, the spatial clustering patterns of high-high aggregation and low-high aggregation, concentrated in the western and southern regions, were detected. 4)The spatial distribution pattern of grassland was affected mainly by natural factors, among which elevation was the dominant factor, explaining up to 42.9% of data variation for grassland spatial distribution. The power of elevation to explain grassland spatial distribution pattern was enhanced by considering other factors as statistically interacting with elevation, in particular livestock industry output, mean annual temperature, density of population and GDP. Under the overall distribution pattern of grassland dominated by elevation, the differences and changes in social and economic factors between regions significantly affected the evolution of grassland spatial distribution patterns, and regional policies also played an important guiding role.