Sex-ratio theory as applied to social Hymenoptera has implicated a large number of factors that can affect predictions about sexual allocation. These factors were examined for the tiny forest ant Leptothorax longispinosus by applying Sewall Wright's path analysis to 10 yr of data. Populations in Vermont and New York were similar with respect to nest sizes and the patterns of queen number but nests in Vermont reared smaller sexuals and produced broods that were more male-biased than their counterparts in New York. Path analysis allowed the exclusion of differential rearing costs, local resource competition, and local mate competition; furthermore, it implicated conflict between queens and workers over alate masses, total investment in sexuals, and allocation to females. Variation residual after path analysis showed that unmeasured parameters were important for reproduction patterns in this species. The most likely unmeasured determinants were resource-acquisition ability, not adequately estimated by worker number, and investment in colony growth. -from Author