AbstractInNewZealand andAustralia, rural landowners believe that local predator control to protect indigenous biota exacerbatesEuropean rabbitOryctolagus cuniculusproblems on their land. We assess the validity of their concerns by reviewing the published literature on effects of predators on rabbit abundance.InNewZealand, where rabbits and their predators are introduced, predators appear to have relatively little effect on rabbit numbers compared with other factors leading to mortality, such as disease, flooding of burrows and burrow collapse. Similarly, inAustralia, rabbit numbers are driven primarily by climate and its effects on food abundance and quality, and by disease. However, where rabbit numbers are low following drought or major epizootics, predation can limit population recovery. In theIberianPeninsula, where rabbits and their predators are indigenous, the effects of predators are unknown, as they are often confounded by other factors. Rabbit numbers are influenced mostly by habitat, food, disease and rainfall. Elsewhere inEurope, predators have their strongest effect when rabbit numbers have been reduced by other factors, but have little effect on high‐density rabbit populations.InAustralasia, abundance of predators (especially rabbit specialists) can usually be predicted from rabbit abundance, not vice versa. Although predation effects can be limiting under certain conditions, they are minor compared to the roles of climate, food, disease and habitat.A key unresolved question is whether those circumstances where predator control might lead to increases in rabbit populations can be identified with enough certainty to allow reliable predictions to be generated. One approach is to implement robust rabbit, predator and disease monitoring programmes at sites with predator control operations. Data on changes in rabbits, predators, and disease prevalence could be combined with local data on other key factors to facilitate reasonable inference about effects of predators on rabbits. The inclusion of carefully matched non‐treatment areas is crucial if such programmes are to succeed.