Using a tractable study system, starfish movements were simulated among rock pools based on empirical data. The authors showed that site fidelity could be a response of animal site selection or the lack of available habitats within the landscape. ‘This dataset accompanies an article in the Journal of Animal Ecology’. ; 1. Understanding site fidelity is important in animal ecology, but evidence is lacking that this behaviour is due to an animal choosing a specific location. To discern site selection behaviour it is necessary to consider the spatial distribution of habitats that animals can occupy within a landscape. Tracking animals and defining clear habitat boundaries, however, is often difficult. 2. We use in situ habitat distribution data and animal movement simulations to investigate behavioural choice in site fidelity patterns. We resolved the difficulty of gathering data by working with intertidal rock pool systems, which are of manageable size and where boundaries are easy to define. Movements of the intertidal starfish Parvulastra exigua were quantified to test the hypotheses that (1) this species displays fidelity to a particular rock pool and that (2) rock pool fidelity is due to site selection behaviour. Observed patterns of individuals (n=10 starfish) returning to a previously occupied rock pool (n = 5 pools per location) were tested against an expected null distribution generated through simulations of random movements within their natural patchy environment. 3. Starfish exhibited site selection behaviour at only one location even though site fidelity was high (av. 7.4 starfish out of 10 found in test pools) in 2 of the 3 locations. The random chance of a starfish returning to a pool increased 67 % for each metre further a rock pool was from the original pool, and 120 % for each square metre increase in surface area of an original pool. The decision of returning to an original rock pool was influenced by food availability. When microalgal cover was > 60 %, there was a ~ 50 % chance of animals staying faithful to that pool. 4. Our results show the importance to consider spatial distribution of habitats in understanding patterns of animal movement associated with animal choices and site fidelity. Returning to a particular place does not necessarily mean that an animal is homing; it may be the only place to go. ; Research Papers and Publications. Science