In order to understand how the apparent freedom of individual movements can concur with social order in a colony, we investigated spatial relationships between foundresses, workers and immature brood in a paper wasp. This is the first time an ecological analytical approach (home range analysis) has been employed to describe small scale spatial use in a social context. In this study, home range was not used in its strictly ecological sense, but rather applied to the comb. Our results show that the positions individual wasps assumed on Polistes dominulus combs are spatially structured. Workers and foundresses do not occupy the comb in random distribution, but rather follow predictable spatial patterns. Each active wasp has its own spatial fidelity area, at least over a 1 d observation period, spending the majority of its time within a very small area averaging approx. 12% of the comb but occasionally it may cover an area of up to 50% of the comb surface. Dominant females occupied a significantly smaller area that either subordinate foundresses or workers. Areas patrolled by workers varied in shape and size, with no relation to time spent on the comb, wasp density or position of immature brood. All the wasps clumped around the dominant female, who proved to be the only individual affected by cell content. Nevertheless, although the wasps tended to clump together, average superimposition rates among nest-mates was very low (higher in workers than foundresses), suggesting that the wasps limit each other’s individual spaces.