In recent years, ecological research has suggested several mechanisms by which biodiversity might affect the risk of acquiring infectious diseases (i.e., the decoy, dilution or amplification effects), but the topic remains controversial. While many experimental studies suggest a negative relationship between biodiversity and disease, this relationship is inherently complex, and might be negative, positive or neutral depending on the geographical scale and ecological context. Here, applying a macroecological approach, we look for associations between diversity and disease by comparing the distribution of human schistosomiasis and biogeographical patterns of freshwater snail and mammal species richness in Uganda. We found that the association between estimated snail richness and human infection was best described by a negative correlation in non-spatial bi- and multivariate logistic mixed effect models. However, this association lost significance after the inclusion of a spatial component in a full geostatistical model, highlighting the importance of accounting for spatial correlation to obtain more precise parameter estimates. Furthermore, we found no significant relationships between mammal richness and schistosomiasis risk. We discuss the limitations of the data and methods used to test the decoy hypothesis for schistosomiasis, and highlight key future research directions that can facilitate more powerful tests of the decoy effect in snail-borne infections, at geographical scales that are relevant for public health and conservation.