Since the late 1960s it became clear that a more sustainable protection of people and property from the negative impacts of natural hazards will require a more balanced use of structural and non-structural measures, such as land-use planning and ecosystem-based solutions for disaster risk reduction, also called Eco-DRR. The most prominent example of Eco-DRR in mountainous regions are forests that protect people, settlements and infrastructures against gravitational natural hazards such as avalanches, landslides and hazards related to mountain torrents. The goal of this paper is to provide an overview on the influence of forests on risks induced by natural hazards and the associated challenges and uncertainties concerning risk analysis. Approaches from natural hazard risk are presented, along with recent results from forest research, thereby offering new ways to integrate forests into risk analysis. We discuss the potential effects of forests on the three important hazard components of the risk concept, namely the onset probability, the propagation probability and the intensity, and propose a set of guiding principles for integrating forests into quantitative risk assessment (QRA) for natural hazards. Our focus thereby lies on snow avalanches, rockfalls, floods, landslides, and debris flows. This review shows that existing methods and models for assessing forest effects on natural hazards suffice for integrating forests into QRA. However, they are mostly limited to the stand- or slope-scale, and further efforts are therefore needed to upscale these approaches to a regional level, and account for uncertainties related to forest effects and natural dynamics. Such a dynamic, rather than a static assessment of risk will finally allow for planning and implementing intelligent combinations of Eco-DRR and technical protection measures.