The spatial heterogeneity of ecosystem services relationships and their driving mechanisms have always been the research focus regarding karst ecosystem services. Taking Guizhou Province, a typical karst mountainous area, as an example, this study selected three ecosystem services:net primary productivity (NPP), soil conservation (SC), and water yield (WY). Based on the accurate estimation of three ecosystem services through the Carnegie-Ames-Stanford approach (CASA) model, the Integrated Valuation of Ecosystem Services and Trade-offs (InVEST) model and the water balance formula, the spatio-temporal change and differentiation characteristics of ecosystem services relationships were analyzed using correlation analysis and root mean square error (RMSE). The driving mechanisms of changes in ecosystem services relationships were then analyzed using a geographic detector modeling. The results showed that:① from 2000 to 2015, the average values of NPP, SC, and WY in Guizhou Province showed an increasing trend. In terms of spatial distribution, NPP and SC presented an increasing trend in the northeast and northwest corner of the study area. The WY increased in the southeastern study area. ② During the study period, the relationships of regional ecosystem services were mainly synergistic, but the synergistic relationship was weakening, accompanied by a trade-off trend. So, due attention should be given to the coordination of ecosystem services relationships. ③ The relationship of regional ecosystem services under different conditions had obvious differentiation characteristics. Compared with dolomite, the trade-off intensity between NPP and SC was greater on limestone; the trade-off intensity between SC and NPP was greater on flat karst basins, and the trade-off intensity between WY and NPP, WY and SC was greater on steep karst slot valley (canyons). The trade-off intensity between WY and SC gradually increased with the increase in vegetation coverage. ④ The results of the geographic detector modeling showed that the total afforestation area was the leading factor affecting the trade-off change between NPP and SC. The total precipitation was the main factor restricting the collaboration between NPP and WY. The elevation and per capita gross domestic product (GDP) were the main factors affecting the relationship between SC and WY. Meantime, the explanatory power of the interaction of any two factors was higher than that of the single factor, indicating the influence intensity of the factor was obviously increased after the interaction. In the future, attention should be paid to multi-factors and their interactions. This study provided deeper insights into the mechanisms of the trade-off/synergy changes among ecosystem services in the karst area, and a scientific reference for the effective protection and management of the karst ecosystem.