Purpose: Grasslands are the largest type of terrestrial ecosystem on the earth, providing rich and unique ecosystem services. However, climate change and human activities have triggered a global degradation of grasslands, which has become a major ecological crisis. In this study, a scientometric analysis was performed to explore the hotspots and frontiers of global grassland degradation research. Materials and methods: Two methods involving visualization were used to analyze these data: document co-citation analysis and burst analysis based on the papers indexed in the Web of Science (WOS) during 1970–2020. Results and discussion: A total of 3580 research papers related to grassland degradation research and 54,666 references were included. The results showed that Harris’s paper in 2010 had the strongest burst value of 26.2, far larger than any other, which shows that this paper was a turning point in the research process. The document co-citation network was divided into 14 main theme clusters. The most influential and emerging research theme clusters were including alpine meadow, grazing exclusion, alpine region, and human activities. Alpine meadow was the largest cluster lasting from 2010 to 2020, indicating that this topic is still active in grassland degradation research. Furthermore, research focus has transferred toward grasslands in Qinghai-Tibetan Plateau. The topic of grazing exclusion is both classic and currently active as it lasted as a research hotspot for 15 years (2004–2018). However, the extent and state of grazing effects research are unclear. Conclusions: As the first scientometric review on grassland degradation research, our study identified the research hotspots and their shifts over the past 50 years, pointing to some potential research frontiers in the future. The scientometric analysis is a useful tool for a quantitative evaluation of research hotspots and trends of global grassland degradation. ; Full Text