Most rock fall hazard zoning methodologies are currently based on trajectory modelling, usually performed along 2D slope profiles. For many topographic configurations, this approach cannot provide a realistic description of the way rock fall trajectories and, ultimately, hazard are spatially distributed all over a slope. This paper presents a new methodology for rock fall hazard zoning, directly applicable to 3D topographies, starting from 3D trajectory simulation results. The procedure is an extension of the Cadanav methodology introduced for hazard zoning along 2D slope profiles. As such, it is fully quantitative and attempts at reducing as much as possible uncertainties and subjective elements affecting current methods for rock fall hazard analysis and zoning. It is also among the first to introduce a “fully-coupled” rock fall intensity-frequency approach. Hazard is estimated by means of “hazard curves”, described at each point of the slope by rock fall intensity-return period couples. These curves may be superimposed on any intensity-return period diagram prescribed in national or regional land use planning regulations, in order to determine which hazardous condition prevails at each point of the slope. The application of the new Cadanav methodology is illustrated for both a theoretical case of simple topography underlying a linear cliff and a real configuration involving a complex topography, characterised by strong three-dimensional features affecting the paths of the blocks. For all topographic models, results obtained for several scenarios involving either localised or diffuse source areas proved that the methodology performs extremely well, providing objective and reproducible results based on a rigorous combination of rock fall energy and return period. Additional tests and real case studies are currently under investigation, for strengthening even further the validation of the approach and extend its applicability to even more complex rock fall scenarios.