Certain animal groups (for example, bats and rodents) are hypothesized to disproportionately harbour and transmit viruses to humans, but whether so-called ‘special reservoirs’ exist is controversial. Mollentzea and Streicker performed an analysis of the literature to construct a dataset of virus–reservoir relationships comprising avian and mammalian reservoir hosts of 415 viruses and the incidence of human zoonotic infections. They found little evidence to support the notion that reservoir host affects the propensity for a virus to infect humans. Rather, the frequency of zoonotic human infections increases with the number of viruses found in each animal order, which is related to the number of species within each group. Their analysis suggests that certain ecological or intrinsic traits of animal groups are unlikely to increase the chance of a virus being a threat to humans.