1. The evolution of hibernation and daily torpor in mammals and birds remains a controversial subject. The original view was that use of torpor reflects a primitive thermoregulation, as it occurs in ancestral groups of mammals. 2. This view is no longer widely supported. However, the interpretation of a polyphyletic derivation of torpor also has been challenged because of the astonishing similarity of torpor patterns among various orders and even the two classes. 3. A recent argument is that mutations required for torpor and hibernation are unlikely to occur simultaneously and that torpor must be plesiomorphic (ancestral), although it is not functionally primitive. Homeothermy is interpreted as a loss of the ability to enter torpor in those groups that could survive without the requirement of heterothermic periods for energy conservation. 4. Interestingly, while torpor in mammals occurs in the phylogenetically old groups, lending support to the hypothesis of an ancestral derivation of torpor, the opposite is the case for birds. Modern bird groups and ancestral mammal groups contain mainly small species that often rely on fluctuating food supply, whereas modern mammalian orders and ancient bird orders contain the largest species with low energy requirements for maintenance and thermoregulation. 5. It is, therefore, possible that not phylogenetic position but size and diet determine the occurrence of heterothermy. Moreover, because endothermy and torpor in birds has apparently evolved separately from that in mammals and because it is possible that daily torpor and hibernation represent two distinct torpor patterns that evolved separately, a convergent evolution of torpor in endotherms cannot be excluded.