Abstract Understanding the behaviour of uranium is essential when assessing the safety of a spent nuclear fuel repository. The geochemical behaviour of uranium, including its reactive transport chemistry, is also a matter of concern when assessing the environmental impact of uranium mining. Subsurface uranium mobility is believed to be primarily controlled by dissolution and (co)-precipitation of uranium mineral solids and adsorption to mineral surfaces. This paper describes a modelling exercise based on characterisation of samples taken from drilled cores at the uranium mineralization at Askola, Southern Finland. In the modelling exercise, current conditions are assumed to be oxidizing and saturated with groundwater. PHREEQC was used for modelling in conjunction with the Lawrence Livermore National Laboratory database, chosen for its extensive coverage of uranium species and mineral phases. It is postulated that weathering processes near the surface have led to uranium dissolution from the primary ore, leaching out from the matrix and migrating along water-conducting fractures with subsequent re-diffusion into the rock matrix. Electron microscopy studies show that precipitated uranium occupies intra-granular fractures in feldspars and quartz. In addition, secondary uranium was found to be distributed within goethite nodules as well as around the margins of iron-containing minerals in the form of silicate and phosphate precipitates. Equilibrium modelling calculations predict that uranium would be precipitated as uranyl silicates, most likely soddyite and uranophane, in the prevailing chemical conditions beneath Lakeakallio hill.