Socioecological models suggest competition for food, foraging efficiency, predation, infanticide risk, and the costs of dispersal regulate primate social structure and organization. Wild populations of squirrel monkeys (Saimiri spp.) appear to conform to the predictions of the predation/competition socioecological model (Sterck et al. in Behav Ecol Sociobiol 41:291-309, 1997) and the dispersal/foraging efficiency model (Isbell in Kinship and behavior in primates. Oxford University, New York, pp 71-108, 2004). However, squirrel monkeys in captivity are reported to maintain patterns of social behavior observed in their wild conspecifics despite different food distribution, predation risk, and dispersal options. This behavioral similarity suggests squirrel monkeys' social behavior has limited flexibility to respond to environmental changes. In this study, we experimentally evaluated the flexibility of social behavior within a captive group of S. sciureus. First, we determined whether dominance and affiliative relationships observed under normal laboratory conditions (with abundant, widely distributed, food; no dispersal option; and no predators) better matched published reports of relationships among wild conspecifics or the predictions of the predation/competition model. Second, we made preferred food items defensible to determine whether dominance interactions would become more frequent and linear, as predicted by the model. The model correctly predicted rates of dominance behavior in both conditions and a linear hierarchy in the defensible food condition but did not predict the consistent affiliative relationships and linear dominance hierarchy observed in normal lab conditions. Although hierarchies were linear and male dominant, manipulating food distribution changed the dominant individual within each sex. Our findings suggest interaction rates adapt more rapidly than social structure to environmental changes in Saimiri and recommend caution in interpreting tests of socioecological models.