Abstract: Watershed-scale runoff changes reflect the combined effects of climate and human activities. However, the attribution of runoff evolution in the Yiluo River basin is currently unclear. Based on the Budyko hypothesis, this study team builds a model to analyze the attribution of runoff changes, quantitatively evaluates the sensitivity coefficient of runoff to changes in various environmental factors, and quantitatively reveals the contribution of various factors to runoff changes. The results show that the runoff of the Yiluo River Basin decreased significantly (α=0.05) from 1970 to 2018, a mutation point occurred in 1991 and the runoff decreased by 46.43 mm (34.32%) during the change period. The sensitivity coefficients of runoff changes to precipitation, potential evapotranspiration and underlying surface parameters are 0.42, −0.18 and −78.36, respectively, that is, every 1 mm increase in precipitation will lead to an increase of 0.42 mm in runoff, and an increase of 1 mm in potential evapotranspiration will lead to a decrease of 0.18 mm in runoff. Each increase of 1 unit of the underlying surface parameter will result in a decrease of 78.36 mm in runoff. The contribution rates of the climate change and human activities to runoff change are 42.60% and 56.36%, respectively, and underlying surface changes caused by human activities are the main reason of runoff changes, among which vegetation change is the dominant factor in underlying surface changes. The research results can provide theoretical support for the attribution of runoff evolution and water resources management planning in the basin.