In a population exhibiting partial migration (i.e. migration and residency tactics occur in the same population), the mechanisms underlying the tactical choice are still unclear. Empirical studies have highlighted a variety of factors that could influence the coexistence of resident and migratory individuals, with growth and body size considered to be key factors in the decision to migrate. Most studies suffer from at least one of the two following caveats: (1) survival and capture probabilities are not taken into account in the data analysis, and (2) body size is often used as a proxy for individual growth. We performed a capture–mark–recapture experiment to study partial migration among juvenile brown trout Salmo trutta at the end of their first year, when a portion of the population emigrate from the natal stream while others choose residency tactic. Bayesian multistate capture–recapture models accounting for survival and recaptures probabilities were used to investigate the relative role of body size and individual growth on survival and migration probabilities. Our results show that, despite an apparent effect of both size and growth on migration, growth is the better integrative parameter and acts directly on migration probability whereas body size acts more strongly on survival. Consequently, we recommend caution if size is used as a proxy for growth when studying the factors that drive partial migration in juvenile salmonid species.