Parasites often exploit more than one host species at any stage in their life-cycle, but the extent to which these host species are used varies greatly. Parasites typically achieve their highest prevalence, intensity and/or abundance in one host species (the principal host), whereas infection levels in auxiliary hosts range from relatively high to very low. The present study examines what influences the distribution of parasite individuals among their different host species, using metazoan parasites that use freshwater fish as their definitive or only host. Specifically, I test the hypothesis that differences in relative infection levels by a parasite among its auxiliary hosts are proportional to the taxonomic distance between the respective auxiliary hosts and the parasite's principal host. Taxonomic distance among hosts is a surrogate measure of their similarity in terms of ecology, physiology and immunology. Using data on 29 parasite species and 6 fish communities, for a total of 47 parasite-locality combinations, it was found that taxonomic distance between the auxiliary hosts and the principal host had no real influence on infection levels in auxiliary hosts, measured as either prevalence, intensity or abundance. The analysis revealed differences in the degree of specialization among major groups of parasites: in terms of abundance or intensity, auxiliary hosts were less important for cestodes than for nematodes and copepods. The lack of an effect of taxonomic distance may indicate that ecological similarity among host species, arising from convergence and not from relatedness, is more important than host phylogeny or taxonomy. Although the results are based on a limited number of parasite taxa, they suggest that parasites may be opportunistic in their colonization of new hosts, and not severely constrained by evolutionary baggage.