Nest predation is an important determinant of owl breeding success. We studied Long‐eared Owl Asio otus productivity and attributes of nest‐sites at the microhabitat and landscape scales in a Mediterranean locality over an 8‐year period. We examined the effect on nest location and productivity of protective cover in concealing the nest from aerial and terrestrial predators. A dense cover of ivy and tree‐foliage at canopy level favoured nest location but not productivity. By contrast, high shrub cover beneath the nest was selected by Owls and was positively related to both the site reoccupancy rate and the overall number of young fledged. Pre‐fledging Owls use the ground, where they are exposed to terrestrial predators, which are much more abundant in the study area than are aerial predators. Our results therefore support the hypothesis that Owls adapt nest‐site choice to local sources of predation risk. As reported elsewhere, Long‐eared Owls in our study area showed restricted territoriality and nested in clusters. As active nest‐sites during the same breeding season were more than 1 km apart on average, and their productivity was never greater for clustered nests than for more isolated nests, nest aggregation could not be interpreted as a case of facultative colonial breeding, which has been reported for this species in other areas. Neither landscape variables indicative of the availability of foraging areas nor structural attributes that protect young from predators explained the remarkable scarcity of nests in half of the study area. Unmeasured factors such as human disturbance could explain the pattern of distribution of Long‐eared Owl nests.