<p>Rainfall-induced landslides pose a significant threat to human life and infrastructure, causing tremendous loss and serious disruption each Year. In this review, we comprehensively examine the development and state of the art in rainfall-induced landslide risk assessment over the past two decades and identify the future needs in this field. The review first introduces a dataset of 560 peer-reviewed articles published on rainfall-induced landslide risk assessment from 2003 to 2022. The existing models of the hazard and vulnerability to rainfall-induced landslides are then classified into four types and the modelling approaches at different scales or elements at risk are discussed. According to the analysis procedures and key characteristics for each type of model, different assessment approaches are also presented and compared. Results reveal a significant increase in the use of machine learning and physical models for hazard assessment, while indicator-weight and data-driven models dominate vulnerability assessment. However, the utilization of different method types at various scales or element types for hazard and vulnerability assessment is not clear. Despite a growing number of studies on rainfall-induced landslide risk assessment in recent years, effectively quantifying the risk from hazard and vulnerability remains challenging. The review concludes by providing a concise summary and proposes prospective research works on rainfall-induced landslide risk assessment.</p>