Aim: To test whether the distribution of the mistletoe Phoradendron californicum is mediated by host distributions (host niche hypothesis, HNH), or by factors such as the mistletoe’s autecology (parasite niche hypothesis, PNH) or that of its vectors (vector niche hypothesis, VNH). Our null hypothesis is that the ecological niche of the mistletoe will not be distinct from that of its hosts or vectors; alternatively, mistletoe infections might appear in hosts only in regions where host distributions overlap suitable conditions for the parasite. Location: Sonoran and Mojave deserts, western North America. Methods: We used ecological niche modelling approaches to summarize suitable environmental conditions for hosts infected and uninfected with mistletoes, as well as for avian dispersers during winter and throughout the year. We compared ecological niches among pairs of species using background similarity tests in relation to the climatic conditions available and accessible to each species. Result:s Niche comparisons supported all PNH expectations, but none of the predictions of HNH or VNH. Main conclusions: Hosts and dispersers of mistletoes generally have distinct ecological niches; mistletoe infections occur in non-random environmental subsets of host and disperser ecological niches; mistletoe infections in different hosts, however, occur under similar climatic conditions. Hence, in this system, the parasite has a rather strictly circumscribed ecological niche, and host species become infected with mistletoe only where they overlap its suitable areas.