||Efforts to investigate crowding phenomena using a population density paradigm, which space or number of organisms or both are varied, have often produced confusing, irregular, or inconsistent results. In addition the experimental manipulations dictated by a density paradigm may be irredeemably confounded. In an attempt to circumvent some of the problems resulting from the use of the density paradigm, an overpopulation paradigm, in which crowding is defined in terms of a ratio between organisms and resources, was employed. For an initial experiment utilizing this paradigm, food was selected as the most suitable resource to manipulate; therefore, the effects of food deprivation on the behavior of group-housed deermice were investigated. Thirty-two male deermice housed in groups of four were presented, in weekly alternation, ad lib. food or food sufficient to maintain the group weight at 80 percent of its free-feeding weight. In addition, for half of the groups the feeding apparatus was constructed such that the mice would have ready access to each other while fedding (social feeding) ; for the other half, the feeding apparatus was constructed such that access to one another while feeding was rendered difficult (isolate feeding). Frequencey of contact and frequency of agonistic behavior were monitored behavior were monitored observationally. Contrary to expectation, the mice displayed significantly less agonistic behavior when deprived than when receiving ad lib. food. Feeding condition (social vs. isolate) had no effect on agonism. neither manipulation affected contact. Neither manipulation affected contact. The reduction in agonistic behavior during deprivation was interpreted in terms of natural poplation-regulating mechanisms. In particular, it was suggested that resource reduction and social dominance factors interact to influence dispersal, which in turn helps regulate population. In addition, it was suggested that the magnitude of effect produced by the food manipulation indicated that the overpopulation paradigm may be a more powerful, as well as less ambiguous, approach to the investigation of crowding.