Abstract Bioenergetics were used to model the potential impacts on adult Atlantic salmon, Salmo salar L., stocks and rod fisheries of removing harbour seals, Phoca vitulina L., from three rivers of different scales in the Moray Firth, Scotland, viz: the Spey (large), Conon (medium) and Moriston (small). Overall, seals had the greatest impact on the Moriston, where removal of a single animal could increase cumulative catch by 17% during the fishing season. On the Conon and Spey the impacts were negligible, and resulted in increased catches of <1% annually. On all rivers eliminating seal predation had the greatest impact during the spring due to the smaller size of spring salmon sub-stocks. A generalised model of seal removal illustrated that stocks and catches increased by >33% in rivers with monthly rod catches ≤10 fish, but declined to <10% for rivers with catches >34 fish. The outputs of the models are qualitative, but provide a management tool for targeting action to resolve seal-salmon fishery conflict. Smaller salmon population units, and spring salmon sub-stocks and fisheries in particular, are most vulnerable to predation. The merit of this approach is discussed regarding the management of Special Areas of Conservation for salmon and seals.