Interspecific interactions play an important role in the success of introduced species. For example, the ‘enemy release’ hypothesis posits that introduced species become invasive because they escape top–down regulation by natural enemies while the ‘invasional meltdown’ hypothesis posits that invasions may be facilitated by synergistic interactions between introduced species. Here, we explore how facilitation and enemy release interact to moderate the potential effect of a large category of positive interactions – protection mutualisms. We use the interactions between an introduced plant (Japanese knotweed Fallopia japonica), an introduced herbivore (Japanese beetle Popillia japonica), an introduced ant (European red ant Myrmica rubra), and native ants and herbivores in riparian zones of the northeastern United States as a model system. Japanese knotweed produces sugary extrafloral nectar that is attractive to ants, and we show that both sugar reward production and ant attendance increase when plants experience a level of leaf damage that is typical in the plants’ native range. Using manipulative experiments at six sites, we demonstrate low levels of ant patrolling, little effect of ants on herbivory rates, and low herbivore pressure during midsummer. Herbivory rates and the capacity of ants to protect plants (as evidenced by effects of ant exclusion) increased significantly when plants were exposed to introduced Japanese beetles that attack plants in the late summer. Beetles were also associated with greater on-plant foraging by ants, and among-plant differences in ant-foraging were correlated with the magnitude of damage inflicted on plants by the beetles. Last, we found that sites occupied by introduced M. rubra ants almost invariably included Japanese knotweed. Thus, underlying variation in the spatiotemporal distribution of the introduced herbivore influences the provision of benefits to the introduced plant and to the introduced ant. More specifically, the presence of the introduced herbivore converts an otherwise weak interaction between two introduced species into a reciprocally beneficial mutualism. Because the prospects for facilitation are linked to the prospects for enemy release in protection mutualisms, species introductions can have complex effects on existing species interactions, between both native and introduced species.