Along the coast of what has come to be known as British Columbia, First Nations face persistent challenges related to the state of the fisheries on which they depend. Fisheries management strategies imposed by the colonial-through-to-federal governance regimes have been implicated in contributing to the challenges, and are rejected by many coastal First Nations who are reasserting governance authority over their fisheries. In particular, the current management approach continues to set ceilings on First Nations' harvest rates (e.g., food, social, and ceremonial allocation). Too often the evidence used to determine such ceilings reflects diets and fishing practices deeply disrupted by social-ecological change, including, but not limited to, colonialism and climate change. Through this paper we use the example of salmon to propose harvest rates more consistent with less disrupted diets, what we refer to as restorative diets. Methodologically, we use empirical records on historical diets as a basis for envisioning what restorative diets might look like and for considering the magnitude of the difference between harvest rates consistent with such diets compared to contemporary diets. We do so by developing a model of restorative harvest rates in reference to caloric needs, the proportion of diets historically contributed by salmon, and the amount of salmon harvested per calorie, which we parameterize using existing empirical records. These methods yield coast-wide restorative harvest rates that range from 68 to 235 kg of salmon per person per year. Such estimates are three to 14 times higher than contemporary rates. We offer the methodology and findings presented here as both catalyst and guidance for further investigations of the conditions (ecological, social, and political) necessary to support the efforts of coastal First Nations, and Indigenous Peoples globally, to restore their fisheries, diets, and food systems.