ABSTRACTStudies on habitat–performance relationships that require joint data on fitness and habitat use are still scarce in long‐lived species. Using data from a southern French population of Mediterranean mouflon (Ovis gmelini musimon × Ovis sp.), we proposed an original approach for gaining information on this relationship by combining a fitness proxy (i.e., carcass mass) collected on harvested rams (n = 257) with knowledge on habitat use obtained from other rams (n = 13) fitted with Global Positioning System (GPS) collars. We first evaluated habitat characteristics encountered by harvested animals in hypothesized home ranges corresponding to circles centered on harvest locations. We set circle size to equal an average ram home range. We found that the carcass mass of harvested individuals decreased with aspect diversity (−16.0% from home ranges with the lowest to the highest diversity), mean slope (−9.3% between flat home ranges and steep ones), and decreasing abundance of open areas (−11.3% between the most and the least open areas). We then tested the robustness of our results by simulating circles with variable sizes and whose centers were randomly located around each harvest location. We found similar results confirming that some habitat characteristics that may be related to resource abundance and spatial structure were important drivers of ram carcass mass in this population. Finally, we showed that simulated circles of variable sizes and centered on GPS locations captured well the habitat composition of home ranges of GPS‐collared rams. Combining different sources of information could hence allow drawing robust inference on key habitats in terms of performance, which is of particular interest when including a spatial component in wildlife management and conservation plans and deciding on appropriate habitat improvements. © 2014 The Wildlife Society.