Habitat quality is an important representation and reflection of the level of ecosystem services, as well as the key link in guaranteeing regional ecological security and improving human welfare. Assessment of the temporal and spatial changes in habitat quality caused by land use change can provide a scientific basis for a region's ecological protection and land management, especially in ecologically fragile areas. For a long time, the Loess Hilly Region has been a typical ecologically vulnerable area in China, and since the 1990s the social and economic conditions of Chinese counties have undergone rapid development. Considering these factors, we used a typical Loess Hilly Region county, Yuzhong County in Gansu province, as an example, and based on ArcGIS platform and the InVEST model, using the land use data of Yuzhong County in 1995, 2005, and 2015, studied for the first time the changes in land use, landscape patterns, and habitat quality from 1995 to 2015; we then analyzed the response of habitat quality to land use change and landscape pattern change, using hot spot analysis and LISA clustering analysis to study and identify the spatial distribution, clustering type, and evolution characteristics of habitat quality of the Loess Hilly Region in 1995, 2005, and 2015. The main conclusions are the following:1) During the study period, both the land use type and landscape pattern changed obviously, and the main land use changes in the study area were that cultivated land was converted to construction land and forest, and there was mutual conversion between grassland and cultivated land; these changes led to an improvement in landscape heterogeneity and the degree of fragmentation; 2) The spatial distribution of habitat quality was:high in the southern areas, medium in the middle areas, and low in the northern areas. During the study period, the area in which habitat quality improved was mainly in the north of the county because of the implementation of the "Grain for Green Project" and the "Ecological Beishan" project; the area of deterioration was primarily located in the center of the county, mainly due to the expansion of construction land, which occupied the farmland and led to an increased degree of fragmentation and a decline in connectivity; 3) The hot spot areas and areas of "High-High" habitat quality type clusters were mainly located in the southern mountain area, while the cold spot areas and areas of "Low-Low" habitat quality type clusters were mainly located in the Loess Hilly Region in the north. Compared with 1995, the degree of habitat quality clustering in 2015 was enhanced:the areas of both the "High-High" type clustering and the "Low-Low" type clustering increased from 1995; the "High-High" type clustering area increasing by 0.68% and the "Low-Low" type clustering area increasing by 0.67%. The paper also puts forward policy suggestions for the optimization of habitat quality in similar areas.