This paper presents a survey of existing methods for motion prediction and risk assessment for intelligent vehicles. The authors classify motion models based on the semantics used to define motion and risk. They discuss the trade-off between model completeness and real-time constraints, and note that the choice of a risk assessment method is influenced by the selected motion model. The paper is divided into three main sections: Physics-based motion models, Maneuver-based motion models, and Interaction-aware motion models. Physics-based models are the simplest, assuming that vehicle motion is governed by physical laws. Maneuver-based models consider that future motion depends on the driver's intended maneuvers. Interaction-aware models take into account the interdependencies between vehicles' maneuvers. The paper also discusses risk assessment, which involves applying these motion models to estimate risk. It classifies existing approaches into two broad families: those that consider only physical collisions between entities, and those that introduce the idea that risk is related to vehicles behaving differently from what is expected given the context. The paper reviews various motion models, including physics-based, maneuver-based, and interaction-aware models, and discusses their limitations. It also presents different approaches to risk assessment, including collision-based and unexpected behavior-based methods. The authors conclude that motion prediction and risk assessment techniques for intelligent vehicles are challenging, as they require reasoning at a high level about a set of interacting maneuvering entities, taking into account uncertainties associated with the data and the models. They suggest that major improvements in this field will be brought by approaches that address both vehicle motion modeling and risk estimation jointly.This paper presents a survey of existing methods for motion prediction and risk assessment for intelligent vehicles. The authors classify motion models based on the semantics used to define motion and risk. They discuss the trade-off between model completeness and real-time constraints, and note that the choice of a risk assessment method is influenced by the selected motion model. The paper is divided into three main sections: Physics-based motion models, Maneuver-based motion models, and Interaction-aware motion models. Physics-based models are the simplest, assuming that vehicle motion is governed by physical laws. Maneuver-based models consider that future motion depends on the driver's intended maneuvers. Interaction-aware models take into account the interdependencies between vehicles' maneuvers. The paper also discusses risk assessment, which involves applying these motion models to estimate risk. It classifies existing approaches into two broad families: those that consider only physical collisions between entities, and those that introduce the idea that risk is related to vehicles behaving differently from what is expected given the context. The paper reviews various motion models, including physics-based, maneuver-based, and interaction-aware models, and discusses their limitations. It also presents different approaches to risk assessment, including collision-based and unexpected behavior-based methods. The authors conclude that motion prediction and risk assessment techniques for intelligent vehicles are challenging, as they require reasoning at a high level about a set of interacting maneuvering entities, taking into account uncertainties associated with the data and the models. They suggest that major improvements in this field will be brought by approaches that address both vehicle motion modeling and risk estimation jointly.