Reproductive success and its determinants are difficult to infer for wild populations of species with no parental care where behavioral observations are difficult or impossible. In this study, we characterized the breeding system and provide estimates of individual reproductive success under natural conditions for a stream-resident, semi-isolated brown trout (Salmo trutta) population. We inferred parentage using a full probability Bayesian model that combines genetic (microsatellite) with phenotypic data. We had tried to exhaustively sample all individuals from a population, including large sib-ship families from three consecutive offspring cohorts. This allowed us to make inferences about the parental genotypes that had produced these families, and thus to augment the parent file with the inferred parental genotypes in cases where large families had unsampled parents. We observed both polygamous and monogamous matings and large reproductive skew for both sexes, but particularly so in males. We found no impact of individual neutral genetic variation or between-partners genetic similarity on reproductive success. Combining parentage analysis with sib-ship reconstruction allows for more precise inference on variance of family sizes and more definite statements on the determinants of reproductive success.