Knowledge of soil erodibility is an essential requirement for erosion prediction, conservation planning, and the assessment of sediment related environmental effects of watershed agricultural practices. This paper reviews the status of soil erodibility evaluations and determinations based on 80 years of upland area erosion research mainly in China and the USA. The review synthesizes the general research progress made by discussing the basic concepts of erodibility and its evaluation, determination, and prediction as well as knowledge of its spatio-temporal variations. The authors found that soil erodibility is often inappropriately or inaccurately applied in describing soil loss caused by different soil erosion component processes and mechanisms. Soil erodibility indicators were related to intrinsic soil properties and exogenic erosional forces, measurements, and calculations. The present review describes major needs including: (1) improved definition of erodibility, (2) modified erodibility determinations in erosion models, especially for specific geographical locations and in the context of different erosion sub-processes, (3) advanced methodologies for quantifying erodibilities of different soil erosion sub-processes, and (4) a better understanding of the mechanism that causes temporal variations in soil erodibility. The review also provides a more rational basis for future research on soil erodibility and supports predictive modeling of soil erosion processes and the development of improved conservation practices.