Dispersal increases the costs of feeding and predation risk in the new environment and is reported to be biased toward habitats similar to the natal region in some mammals. The benefits and costs of dispersal often differ between sexes, and most mammals show male-biased dispersal in relation to a polygamous mating system. Japanese serow is generally a solitary and monogamous species. However, recent studies have shown that the sociality of serows on Mt. Asama differs between habitat types. In the mountain forests with low forage availability, solitary habits and social monogamy were observed, while, in alpine grasslands, female grouping and social polygyny were observed, which is probably due to abundant forage availability. We investigated the effects of habitat characteristics and sociality on the dispersal of serows using fecal and tissue samples from two different habitats on Mt. Asama. The Fst value between the two areas was significantly positive, and the mean relatedness within areas was significantly higher than that between areas, which suggests limited gene flow and natal habitat-biased dispersal. Bayesian clustering analysis showed unidirectional gene flow from forest to grassland, which was probably due to the high forage availability of the grassland. Analyses of the assignment index and mean relatedness did not show male-biased dispersal, even in the grassland, where serows were polygynous. Thus, polygyny in the grassland is not linked to male-biased dispersal. In summary, our study suggests that dispersal patterns in Japanese serows are affected by habitat rather than social differences.