Geometallurgical model aims to capture geological and metallurgical variation of an ore in a 3D map. This is an important part of production planning and management. In creating such a model geological and metallurgical information is combined but current practices almost solely ignore the liberation information. This paper describes a technique how this important data can be taken into a geometallurgical model. Malmberget, the case study iron ore in Northern Sweden, consists of several ore lenses with variable mineralogy illustrated e.g. in magnetite-hematite ratio. Modal mineralogy does not fully explain the metallurgical response. To include ore texture and mineral liberation factors two sub-models were created based on liberation analyses of metallurgical testing and the Particle Tracking technique. Consequently, for each ore block, the particle breakage model gives the liberation distribution. Thereafter the process model, consisting of unit operations with property based models, forecasts the metallurgical response.