Life history of salmonids is highly flexible, with the most prominent feature being migratory polymorphism within single populations. Understanding the mechanisms for the coexistence of large migratory forms and small resident counterparts provides fundamental insights into evolutionary and ecological processes of diversification. In addition, since large migratory salmonids significantly alter ecosystems by their dead carcases and are also important as human food resources, predicting the frequency of both forms under changing environments has been eagerly required. Past studies have assumed the existence of a threshold point that determines the divergence of resident and migratory tactics: individuals that exceed a threshold body size at a juvenile stage become resident because they can acquire enough food to mature within freshwater. Here, I propose the concept of “relative assessment” in which individuals choose alternative tactics based not only on their own condition but also on relative condition in a given population. For example, individuals should adopt migratory tactic even though they exceed a presumed threshold size when other individuals are larger. In this talk, I will present theoretical and empirical studies on masu salmon and discuss how the idea of relative assessment changes the perspectives of life history divergence and dynamics of alternative tactics.