It is normal for hosts to be coinfected by parasites. Interactions among coinfecting species can have profound consequences, including changing parasite transmission dynamics, altering disease severity, and confounding attempts at parasite control. Despite the importance of coinfection, there is currently no way to predict how different parasite species may interact with one another, nor the consequences of those interactions. Here we demonstrate a method that enables such prediction by identifying two nematode parasite groups based on taxonomy and characteristics of parasitological niche. From an understanding of the interactions between the two defined groups in one host system (wild rabbits), we predict how two different nematode species, from the same defined groups, will interact in coinfections in a different host system (sheep), and then we test this experimentally. We show that as predicted, in coinfections, the blood-feeding nematode Haemonchus contortus suppresses aspects of the sheep immune response, thereby facilitating the establishment and / or survival of the nematode Trichostrongylus colubriformis; and that the T. colubriformis-induced immune response negatively affects H. contortus. This work is the first to use empirical data from one host system to successfully predict the specific outcome of a different coinfection in a second host species. The study therefore takes the first step in defining a practical framework for predicting interspecific parasite interactions in other animal systems.