Although the behaviour of individuals is known to impact the genetic make-up of a population, observed behavioural patterns do not always correspond to patterns of genetic structure. In particular, philopatric or dispersal-limited species often display lower-than-expected values of relatedness or inbreeding suggestive of the presence of cryptic migration, dispersal, or mating behaviours. I used a combination of microsatellite and mark–recapture data to test for the influence of such behaviours in a dispersal-limited species, the Pacific jumping mouse, within a semi-isolated population over three seasons. Despite short dispersal distances and a low rate of first generation migrants, heterozygosities were high and inbreeding values were low. Dispersal was male-biased; interestingly however, this pattern was only present when dispersal was considered to include movement away from paternal home range. Not unexpectedly, males were polygynous; notably, some females were also found to be polyandrous, selecting multiple neighbouring mates for their single annual litter. Patterns of genetic structure were consistent with these more inconspicuous behavioural patterns. Females were more closely related than males and isolation by distance was present only in females. Furthermore, detailed genetic landscapes revealed the existence of strong, significant negative correlations, with areas of low genetic distance among females overlapping spatially with areas of high genetic distance among males. These results support the hypothesis that the detected cryptic components of dispersal and mating behaviour are reducing the likelihood of inbreeding in this population through paternally driven spatial mixing of male genotypes and polyandry of females.