The major objective of this study was to test the hypothesis that juvenile Atlantic salmon kin occupy adjacent territories in their natural habitat in order to profit from the benefits associated with kin-biased behaviours, as has been observed under controlled laboratory conditions. Microsatellites were used to establish the relatedness of salmon fry (in their first summer of life) and parr (in their second and third summer of life) captured in adjacent territories. We did not observe a relationship between the proximity and the relatedness of either parr of the same cohort or fry in their natural habitat. Although many pairs of fry were identified as being related when sampled immediately after emergence, most family groups did not occupy adjacent territories. The high dispersal potential in rivers, the low occupation rate of the habitat and the incidence of half-sibs in nature most probably reduce the opportunity and advantage of kin-biased behaviour, in contrast to laboratory studies conducted in artificial, high-density conditions.