Historically, the importance of dispersal ability for the distribution of organisms has often been ignored, partly because of the difficulty of measuring it. Many unionid mussels, which have larvae that are obligate parasites (usually on fish), are endangered and the conservation and management of these mussels depend on knowledge of the main drivers of their distribution. Metacommunity theory predicts that limited dispersal should weaken the association of community composition with environmental factors. We tested this prediction by comparing the strength of association with environmental factors of (i) mussels with different dispersal abilities based on mobility of known host fish and (ii) mussels with different host infection strategies targeting fish with different mobility. Mussels with more mobile host fish showed a significantly stronger association with host fish presence and catchment (as a proximate measure for large-scale differences in environmental conditions or as a spatial component) compared to mussels with less mobile host fish. Our results thus indicate that mussel distribution is more closely linked to host fish for high-dispersal mussels, which suggests the potential for species sorting. Mussel species with weak dispersal capabilities show the opposite pattern and are therefore less able to colonise all suitable sites (species sorting with limited dispersal). Thus, the absence of mussels does not necessarily indicate unsuitable environmental conditions, but can also be caused by dispersal limitation. The impact of different host infection strategies was less clear, and differences in responses could be driven by differences in host fish or even habitat specialisation in addition to potential differences in dispersal abilities. A better knowledge of the dispersal via host fish of unionid mussels and their host infection strategies will be crucial to understanding their metacommunity dynamics, a necessary precursor for effective conservation practices.