Abstract This work evaluated thinning strategies in uneven-aged Norway spruce forest stands using a tree distance dependant growth model. The model is based on competition for light between trees demonstrated through interception of light rays by individual crowns. It was adapted to mountain settings by analyzing the effect of slope and aspect on light distribution. Growth functions were built by empirically comparing height and diameter increments with age and light availability. A maximum number of new trees set by the user was added each year and mortality, i.e. disappearance of trees not receiving enough light to survive was also included. The model is part of the CAPSIS software package which provides a user-friendly interface for silvicultural simulations. First, thinning typical of current practices in uneven-aged Norway spruce stands in the northern French Alps was simulated. This thinning appeared to be neutral in terms of wood production and beneficial but not dynamic enough in terms of stand renewal. In a second experiment, we compared group selection to individual selection in a virtual two-layered stand organized into clumps. This showed that individual selection favors the growth of the largest trees in the stand while group selection favors the growth of the smaller ones as well as regeneration. In a third simulation, we generated a strategy to convert an even-aged stand into an uneven one. Success appeared to be linked to a flexible approach alternating high and low intensity cuttings, group and individual selection.