A variety of definitions involving body temperature (Tb), metabolic rate and behavior have been used to define torpor in mammals and birds. This problem is confounded in some studies of free-ranging animals that employ only skin temperature (Tsk), a measure that approximates but may not precisely reflect Tb. We assess the accuracy of Tsk in the context of a recent definition for torpor called active temperature. We compared the active temperatures of individual big brown bats (Eptesicus fuscus), which aggregate in cavities, with solitary, foliage-roosting hoary bats (Lasiurus cinereus). In captive big brown bats, we compared Tsk and core Tb at a range of ambient temperatures for clustered and solitary roosting animals, compared Tsk and Tb during arousal from torpor, and quantified the effect of flight on warming from torpor. Hoary bats had significantly lower active temperatures than big brown bats despite having the same normothermic Tsk. Tsk was significantly lower than Tb during normothermia but often greater than Tb during torpor. Flight increased the rate of warming from torpor. This effect was more pronounced for Tsk than Tb. This suggests that bats could rely on heat generated by flight muscles to complete the final stages of arousal. Using active temperature to define torpor may underestimate torpor due to ambient cooling of external transmitters or animals leaving roosts while still torpid. Conversely, active temperature may also overestimate shallow torpor use if it is recorded during active arousal when shivering and non-shivering thermogenesis warm external transmitters. Our findings illuminate the need for laboratory studies that quantify the relationship between metabolic rate and Tsk over a range of ambient temperatures.