The complexity of parasite communities is well known because the habitats are composed of live organisms and the life cycles are often complex (e.g. endoparasites often require 1-4 host species). Thus, parasite communities should not be analysed simplistically by using only a few variables and descriptors, because such as approach will provide limited information. The study of these communities can be greatly improved if we analyse them using several methods, descriptors and host attributes. This research compares and analyses all the metazoan parasite species of a fish assemblage (Labridae) from the Great Barrier Reef. These analyses considered six parasitological descriptos, host and parasite ecological characteristics as well as the phylogeny of hosts, and ecto- and endoparasite communities in 2 hierarchical levels. Two hundred and ninety-five fishes belonging to 30 species were collected at Lizard Island, Australia, between 2002 and 2004. All fishes were examined for parasites and approximately 38,000 parasite individuals were collected. I identified 110 parasite species; 33 were ectoparasites, mostly isopods and copepods, and 77 endoparasite species, mainly cestodes. The analyses of parasite infracommunities revealed that host and parasite body sizes were not strong influences on the parasite community structure. There were large differences in parasite body sizes among species although most parasite species were small. Parasitological descriptors (richness, abundance, biomass and densities) were generally not correlated with host body size. There were positive and significant correlations between infra- and component community richness and between richness, abundance and biovolume of parasites. Moreover, the random assortment model (RA) provided a good fit for the whole community of parasites in seven of 14 fish species. This indicates that interspecific competition was not important in structuring these communities. These results revealed that parasite infracommunities were unsaturated with species, therefore there was unlimited resources and no interspecific competition. These observations suggest that parasite communities in wrasses are isolationists. However, highly rich and abundant communities, similar to those from wrasses, have been generally classified as interactive. One possible explanation for these results is that small parasites require only a small amount of resources, and therefore they do not need to compete. The analysis of component communities of parasites revealed that the host phylogeny and ecology were not strong influences on the structure of parasite communities of 14 fish species. Parasite distributions across fish species were incongruent with the host phylogeny, perhaps because there were relatively high number of parasites identified as types as well as several generalist and infrequent parasites. Only some congeneric fish species (Cheilinus chlorourus /C. trilobatus, and Thalassoma jansenii /T. lunare) had similar endoparasite composition, but the other 10 fish species differed significantly in all variables. A few ecological host characteristics affected parasite communities. The correlations between host ecological characteristics and parasitological descriptors controlled for host phylogeny showed that host body weight was determinant for richness and abundance of ectoparasite communities only. I suggest that positive relationships between these descriptors may be due to direct transmission of ectoparasites, because large hosts offer a large area to be colonized by parasites. The indirect transmission of endoparasites from one host to the next may be affected by stochastic factors. Although, larval parasite composition may respond both to host diet and to predator-prey interactions because this is the path by which many parasites complete their life cycles. However, high aggregate distribution of parasites, wide range of invertebrate and vertebrate intermediate hosts, and low predation rates by their predators suggest that parasites of wrasses have great difficulty in reaching the definitive hosts. This study has demonstrated that parasites of wrasses belong to a diverse and complex ecosystem in which only a few variables were influential in the structure of the parasite communities. However, other variables emerge as potential important determinants, for example, body size of parasites that can be analysed jointly with interactive-isolationist community indices, transmission of parasites (e.g. number of hosts in the life cycle) and predator-prey interactions, such as distribution and abundance of predators, and predation rate of wrasses. The importance of these variables and their association with one another needs to be evaluated to understand the processes that structure these parasite communities in wrasses, and in other hosts involved in the life cycles of the parasites.