The increasing length of roads, the deepening of land transport corridors, and the expansion of land transport networks have a dramatic impact on the functioning and integrity of global ecosystems.In this paper, CiteSpace is used to carry out the visual knowledge graph analysis of road ecological impact research. Starting from the aspects of time evolution and hot research, this paper reviews the research status, evaluation methods and potential benefits of road ecological impact research, reveals the current problems in this research, and predicts the research trend in this direction.The research on road ecological impact has experienced three stages: initial stage, rapid development stage and hot spot development stage. From discussing the negative impact of a single road, it has gradually developed into the comprehensive impact of traffic corridor and road network on multiple ecological factors such as water, soil, air and biomass. The hot spots are concentrated in China and the United States. Focus on water system cutting and hydrological changes, soil physicochemical changes and soil erosion, air pollution, carbon emissions and microclimatic effects, individual and population of wild animals, plant communities and coverage, landscape and habitat fragmentation, as well as their effects of time accumulation and space superposition, ecological impact evaluation and benefit analysis.At present, there are many problems in ecological impact analysis, such as more qualitative and less quantitative analysis, weak dynamic assessment, insufficient accuracy, unclear cumulative and overlapping effects of roads, incomplete impact assessment, lack of quantitative methods, unclear benefits and contributions, and insufficient prediction accuracy. Future research should pay more attention to the temporal and spatial scale benefits, cumulative and superimposed effects of roads, the fine quantification of monitoring and evaluation, the determination of benefits and contributions, forecasting and planning, and the protection from the source.