1. Hibernation is the most effective means for energy conservation during winter in mammals. The drawbacks of deep and prolonged torpor include reduced immunocompetence, and consequently, hibernators should be selected to minimize torpor expression when climatic conditions or energy availability (e.g. food or fat stores) permit. Therefore, it seems surprising that some hibernators employ extraordinary long hibernation seasons, lasting well beyond periods with unfavourable conditions. 2. Because of their extended use of torpor, edible dormice (Glis glis) provide an ideal model for scrutinizing interactions between energy reserves (i.e. body fat stores) and thermoregulatory patterns. We used a multimodel inference approach to analyse body temperature data (i.e. use of torpor) from 42 entire hibernation seasons over 4 years in females in relation to body mass. 3. Body mass prior to hibernation did not affect the duration of the hibernation season, but animals hibernated for c. 8 months, that is, 2 months longer than required by environmental conditions. Fatter individuals aroused significantly more often, had a higher mean minimum body temperature during torpor and remained euthermic for longer periods than leaner animals. 4. Surplus energy was therefore not used to shorten the hibernation season, but to rewarm more frequently, and to allow shallower torpor bouts. These adjustments apparently serve to avoid negative effects of torpor and, perhaps equally importantly, to minimize the time active above-ground. We argue that maintaining a short active season, despite surplus energy reserves, may be explained by known beneficial effects of hibernation on survival rates (via predator avoidance). 5. Our data provide quantitative evidence that hibernation is a flexible tool within life-history strategies. We conclude that, apart from energetic necessities due to harsh environmental conditions, predator avoidance may be an important factor influencing patterns of hibernation and torpor in mammals. Thus, our study indicates that climatic conditions alone are not a good predictor of hibernation patterns or survival in hibernating species during global climate change.