Apex predators structure ecosystems by hunting mesopredators and herbivores. Their ecological influence is determined not only by the number of animals they kill, but also by how prey alter their behaviours to reduce risk. Predation risk is variable in space and time creating a landscape of fear. In Australia, dingoes hunt red foxes and suppress their populations. As both predators are commonly subjected to eradication programs, the question arises whether humans alter the risk dingoes pose to foxes and in turn alter the foxes’ avoidance behaviours. We studied the spatio-temporal activity patterns and wariness behaviours of foxes and dingoes at sites where they were protected (predator friendly), where they were persecuted (predator persecuted), and at sites where foxes were persecuted, and dingoes had been eradicated (dingo eradicated). The landscape of fear hypothesis predicts that foxes will be the most spatiotemporally restricted and most fearful at predator friendly sites, and least restricted and fearful at dingo eradicated sites. We found that fox occupancy was highest at dingo eradicated sites; and that they avoided times of heightened dingo activity at predator friendly sites more than at predator persecuted sites. Contrary to predictions, foxes were the least fearful (lowest frequency of cautious and vigilant behaviour) and most social (frequency of social interactions) at predator friendly sites. Our findings suggest that in the absence of persecution, mesopredators living with socially-stable apex predators can anticipate and avoid risk, reducing the need for constant vigilance (i.e., fear). Where predators are protected, predator avoidance may be driven by knowledge rather than fear alone.