Luossavaara-Kiirunavaara AB (LKAB) operates one iron ore mine, three concentration plants and three pelletising plants in Kiruna, Sweden. The current methods of separation at the beneficiation plants are low intensity magnetic separation (LIMS) and reverse apatite flotation, where the wet LIMS stage is regarded as the crucial part of silica separation from the ore. Because future silica levels are expected to increase in the incoming material to the beneficiation plants, geometallurgy has become an important area of development and research.A laboratory-scale methodology was developed for the systematic characterisation of the ore for mineral processing, with the focus on the high silica ore type B2 and newly discovered subtypes. This methodology combines mineralogical, geochemical and process mineralogical characteristics. Automated mineralogy (QEMSCAN) is used to study the modal mineralogy, the distribution of silicate minerals in different particle size classes after comminution, the deportment of silicon (Si) between various silicates and the degree of liberation and intergrowth of magnetite and silicates. An essential stage in the systematic characterisation of the ore deposit for mineral processing is the estimation and simulation of different parameters from the laboratory scale to the full scale at the concentration plants. This is achieved by using empirical models based on the data collected at the mineral processing plants. As the first step to implement the geometallurgical information, the current geological model is reconstructed with the mineralogy and the two, newly defined subtypes of high silica ore type B2 are separated into different domains.To study the reduction of the SiO2 content in the magnetite concentrate, a large-scale reverse cationic silicate flotation laboratory test program was initiated in early spring 2016. This program investigated potential silicate flotation in the beneficiation plants at the Kiirunavaara site. The second target is to implant mineral processing parameters into the block model based on the new geological model. This information provides a good basis for making a prognosis for mine planning and a quality prognosis regarding the crude ore for mineral processing.CITATION:Niiranen, K, 2017. A geometallurgical perspective on a high-grade iron ore deposit at Kiirunavaara, northern Sweden, in Proceedings Iron Ore 2017, pp 309-318 (The Australasian Institute of Mining and Metallurgy: Melbourne).