In their recent study, Beaman et al. [1] measured the metabolic response of the most abundant Arctic seabird, the dovekie (Alle alle), to experimental variation in air temperature. Based on their results on energy and water expenditure in captive conditions, Beaman et al. conclude that this Artic species should not suffer much from direct consequences of climate change. Indeed, their estimated upper critical temperature TUC (i.e. temperature above which metabolic rate increases) was 22.4°C, which is ca. 15°C higher than the average air temperature, and ca. 2°C higher than the maximum temperature currently recorded at the breeding colony. While projecting their data under various climatic scenario, Beaman et al.[1] conclude that dovekies should even benefit of the ongoing climate warming from a thermoregulatory perspective. While the data collection is quite standard in the field of thermal biology [2], I would like to offer my personal view on the potential perils of concluding about wild animal response to climate change based on data collected in ‘a small dark box in the lab’. Classically, the thermal response of animals is evaluated using respirometry, which enables to quantify gas exchange (i.e. O2, CO2, H2O) between an animal and its environment in response to variation in ambient temperature within a small metabolic chamber [2]. This is usually conducted in a post-absorptive state and in the dark to limit energy expenditure linked to digestion and physical activity. While such conditions are highly relevant to evaluate the minimum energy requirements of animals (i.e. basal metabolic rate), I question its relevance for evaluating the thermal responses of wild animals to the heat.