Abstract An enduring challenge for community ecologists is to understand how different assemblages of species are derived from a common pool of potential colonists. Early colonisers can affect the ability of subsequent arrivals to colonise and persist, and thereby influence the development and structure of the resulting assemblage. In two independent experiments, we tested for such historical effects of priority on the assemblages of mobile fauna colonising artificial seagrass initially occupied by epibiota (algae and sessile invertebrates) and/or predatory Palaemonid shrimp. Multivariate analyses detected strong priority effects of both epibiota and Palaemonids on the structure of faunal assemblages as early as 5 days, and up to 45 days after experimental treatments were established, even though initial conditions (i.e. abundances of epibiota and Palaemonids) had become similar among treatments. The abundances of key taxa, identified by SIMPER analyses as those taxa contributing the most to multivariate differences among treatments, were typically enhanced where epibiota were initially present. In contrast, prior colonisation by Palaemonid shrimp produced subtle and variable effects on individual taxa. Nevertheless, these experiments provide evidence for priority effects and demonstrate how variation in the structure of contemporary assemblages can be intimately linked with key historical events that occurred during their development, but may no longer be apparent. Consequently, our ability to interpret variation among contemporary assemblages may be enhanced when contemporary patterns are viewed within an historical context.