Boreal forests are an important source of timber and pulp wood, but provide also other products and services. Utilizing a simulation program and field data from a tree retention experiment in a Scots pine forest in central Sweden, we simulated the consequences during the following 100 years of various levels of retention on production of merchantable wood, dead wood input (as a proxy for biodiversity), and carbon stock changes. At the stand level, wood production decreased with increased retention levels, while dead wood input and carbon stock increased. We also compared 12 scenarios representing a land sharing/land sparing gradient. In each scenario, a constant volume of wood was harvested with a specific level of retention in a 100-ha landscape. The area not needed to reach the defined volume was set-aside during a 100-year rotation period, leading to decreasing area of set-asides with increasing level of retention across the 12 scenarios. Dead wood input was positively affected by the level of tree retention whereas the average carbon stock decreased slightly with increasing level of tree retention. The scenarios will probably vary in how they favor species preferring different substrates. Therefore, we conclude that a larger variation of landscape-level conservation strategies, also including active creation of dead wood, may be an attractive complement to the existing management.