The behavioural dominance hypothesis suggests that differential migration among individuals of a species of bird is due solely to social interactions that force subordinate individuals (often a class, e.g. female or young birds) to migrate farther into the winter range than dominant individuals (often a class, e.g. make or old birds). Here, this hypothesis was tested with two experiments. In the first experiment, the prediction was tested that dominance acts within a sex-age class and influences migration distance. The outcomes of interactions within dyads of dark-eyed juncos, Junco hyemalis, were observed. Each dyad consisted of a junco caught in winter in Michigan USA matched with another of the same sex-age class caught in Indiana, which is situated farther south and therefore farther into the winter range of this species. Michigan birds were dominant in only half of the experimental dyads (21 of 41 dyads), which is inconsistent with the prediction of the dominance model. In the second experiment the prediction was tested that members of a sex-age class that migrates farther into the winter range should be subordinate to members of a different class that migrates a shorter distance. Young males that wintered in Michigan were pitted against old males that wintered in indiana. In 19 of 25 dyads, the more southern-wintering old males were dominant, which also is counter to the prediction of the dominance hypothesis. These results indicate, at the very least, that in migratory J. h. hyemalis, dominance does not play as important a role in determing latitude of winter residence as has been suggested.