We investigated a simulation model of neighborhood competition and compared the predictions of this model to empirical patterns of ant diversity and spatial arrangement. We modeled two kinds of competition: mortality of juveniles that found nests near established colonies and competition between established colonies through local crowding. The results of the model indicated that spatial arrangement is affected by the relative strength of intra- and interspecific competition as well as by colony birth and date rates. Thus regular spatial arrangement of colonies cannot be used as a test of any type of competitive effects, since other spatial arrangements can also result from strong competition between colonies. Regular intraspecific spatial arrangement is most readily produced by increased mortality of young colonies that settle near established conspecifics (no competition between species). In these model communities, spacing is increasingly regular as density increases (density-dependent spacing). Where competition occurs only between established colonies, nest spacing is random, and the spatial statistics are not related to nest density. Persistence was defined as the length of time that all species in a simulation coexisted. Where competition was intense, intra- and interspecifically, persistence in generations was increased by low colony death rates and by large absolute numbers of colonies. The adult-adult component of interspecific competition is relatively less important to species persistence than are adult-juvenile effects.