Many hypotheses attempt to explain parasite–host associations, but rarely are they examined together in a single community. For hosts, key traits are the proportion of infected individuals (prevalence) and the diversity of parasites infecting them. A key parasite trait is host specificity, ranging from specialists infecting one or a few closely related species to generalists infecting many species. We tested 10 hypotheses to explain host-parasite associations; five ‘host-centric’ (e.g. prevalence is related to host abundance) and five ‘parasite-centric’ (e.g. parasite abundance is related to host specificity). We analyzed a community of 67 locally transmitted avian haemosporidian parasite lineages (genera Plasmodium, Haemoproteus or Leucocytozoon), sampled from 2726 birds (64 species) in southern Sweden. Among host-centric hypotheses, Haemoproteus and Leucocytozoon prevalence and Haemoproteus diversity were related to host habitat preferences, whereas there were no relationships with host abundance or body mass. Haemoproteus and Leucocytozoon prevalences were more similar among closely related than among distantly related host species. Haemoproteus prevalence and diversity were lower in host species with few close relatives (‘evolutionarily distinct’ hosts). Among parasite-centric hypotheses, most lineages, even relative generalists, infected closely related host species more often than expected by chance. However, the host species of Haemoproteus and Leucocytozoon lineages overlapped less among lineages than expected by chance. Specialists did not reach higher prevalences than generalists on single host species. However, the abundance of Haemoproteus lineages was related to host specificity with generalists more common than specialists; this was driven by three closely related generalists. Host specificity of parasites was unrelated to the abundance or evolutionarily distinctiveness of their hosts. Parasite communities are likely structured by many factors and cannot be explained by hypotheses focusing solely on hosts or parasites. However, we found consistent effects of host phylogenetic relationships, plausibly a result of evolutionarily conserved host immune systems limiting parasite distributions.