In nature, light is a key driver of animal behaviour and physiology. When studying captive or laboratory animals, researchers usually expose animals to a period of darkness, to mimic night. However, ‘darkness’ is often poorly quantified and its importance is generally underappreciated in animal research. Even small differences in nocturnal light conditions can influence biology. When light levels during the dark phase are not reported accurately, experiments can be impossible to replicate and compare. Furthermore, when nocturnal light levels are unrealistically dark or bright, the research is less ecologically relevant. Such issues are exacerbated by huge differences in the sensitivity of different light meters, which are not always described in study methods. We argue that nocturnal light levels need to be reported clearly and precisely, particularly in studies of animals housed indoors (e.g. ‘<0.03 lux’ rather than ‘0 lux’ or ‘dark’), and that these light levels should reflect conditions that the animal would experience in a natural context.