Using data on bird knowledge from a Zapotec community in northern Oaxaca, this paper addresses a fundamental issue in the theory of ethnobiological classification: to what extent do the ways people group and name biological organisms reflect a “natural” scheme similar to that adopted in scientific taxonomy rather than reflecting other features of an organism's ecology or behavior that are important for local people? We recorded 209 bird species, corresponding to 30 folk generics, 77 folk specifics, and 11 varietals, in the sense used by Berlin et al. (1973). However, Zapotec research subjects assigned birds to four main named groupings that cut across life-forms: nocturnal forms, forms that walk or alight on the ground, forms that fly high in the open sky, and forms that fly through the canopy or through other vegetation. Our data suggest that even if there is psychological evidence to support some separate natural way of classifying birds based primarily on morphology, Zapotec routinely and primarily think of birds in ecological and behavioral contexts from their everyday language and grouping practice.