Long-lived animals with a low annual reproductive output need a long time to recover from population crashes and are, thus, likely to face high extinction risk, if the current global environmental change will increase mortality rates. To aid conservation of those species, knowledge on the variability of mortality rates is essential. Unfortunately, however, individual-based multi-year data sets that are required for that have only rarely been collected for free-ranging long-lived mammals. Here, we used a five-year data set comprising activity data of 1445 RFID-tagged individuals of two long-lived temperate zone bat species, Natterer’s bats (Myotis nattereri) and Daubenton’s bats (Myotis daubentonii), at their joint hibernaculum. Both species are listed as being of high conservation interest by the European Habitats Directive. Applying mixed-effects logistic regression, we explored seasonal survival differences in these two species which differ in foraging strategy and phenology. In both species, survival over the first-winter of an individual’s life was much lower than survival over subsequent winters. Focussing on adults only, seasonal survival patterns were largely consistent with higher winter and lower summer survival but varied in its level across years in both species. Our analyses, furthermore, highlight the importance of species-specific time periods for survival. Daubenton’s bats showed a much stronger difference in survival between the two seasons than Natterer’s bats. In one exceptional winter, the population of Natterer’s bats crashed, while the survival of Daubenton’s bats declined only moderately. While, our results confirm the general seasonal survival pattern typical for hibernating mammals with higher winter than summer survival, they also show that this pattern can be reversed under particular conditions. Overall, our study points towards a high importance of specific time periods for population dynamics and suggests species-, population- and age class-specific responses to global climate change.