Aim: Geographical variations in endotherm body size (e.g., Bergmann’s rule/James's rule and Allen’s rule) have long been tested. However, the patterns and causes of geographical variation in body size within bat species, especially within widespread hibernating species, are little known. Here, we evaluated the possible causes of geographical size variation patterns in the greater horseshoe bat (Rhinolophus ferrumequinum), a bat species widely distributed across the Palearctic. Location: Palearctic Methods: We collected body size data (body mass and forearm length) from 1172 adult bats via direct measurements or from the literature. We used high-resolution environmental data for the sampled sites. We applied multiple linear regressions and an information-theoretic approach on separate female and male datasets to estimate the support for various hypotheses. Results: Temperature and latitude failed to predict body mass in both females and males. The best model showed that populations with longer forearms in both sexes were associated with higher mean temperatures (in the warmest quarter) and lower mean temperatures (in the coldest quarter). The independent contribution of these temperature parameters to forearm length in both sexes was higher than that of all other variables. Main conclusion: The observed relationship between temperature and body mass was not consistent with James’s rule, and thus failed to support the heat conservation hypothesis. Our results are consistent with the predictions of Allen’s rule and indicated that direct heat dissipation from the wing may be the most likely mechanism underlying geographical variation in forearm length. We suggest that the evaluation of the relative importance of multiple causal mechanisms may improve our understanding of patterns of geographical variation in endotherms.