In many temperate zone bat species, large numbers of individuals aggregate in autumn at potential hibernacula such as caves, mines, and cellars, for a behaviour known as swarming. This autumn swarming probably serves two functions, the achievement of matings and the assessment of potential hibernation sites. Even though this remarkable behaviour has attracted a lot of research during the last decades, several central questions regarding the autumn swarming of temperate zone bats remain unanswered. One of them is how juvenile bats (defined here as young of the year) are able to find swarming sites, which are typically dozens of kilometres away from the maternity colonies where they had been born. In this study, we used a combination of field data and population genetic tools to assess whether the juveniles are likely to learn the location of swarming sites from their mothers. To obtain non-destructive wing-tissue samples for DNA-based maternity assignments, 170 Myotis daubentonii and 195 Myotis nattereri were captured during the swarming seasons in two subsequent years, at a large hibernaculum in Germany. Based on 14, respectively 13, highly polymorphic microsatellite loci, maternity assignment tests were conducted for all captured adult females and juveniles that had been born in the two respective years. For M. daubentonii we found four assigned mother-offspring pairs, whereas in M. nattereri, eight mother-offspring pairs could be assigned with high certainty. Moreover, among the latter species, in seven pairs the mothers and the assigned offspring were caught within ten minutes of one another on the same night. Using a simulation, we show that for M. nattereri, significantly more juveniles than expected are caught together with their mothers at the hibernacula. We discuss the implications of our findings with regard to understanding autumn swarming behaviour in temperate zone bats, as well as for the conservation of bats that depend on swarming sites for mating and hibernation.