The efficiency of spatial conservation measures for threatened species depends mostly on the proportion of time that animals spend within the protected areas. We illustrate this with our case study of the population of recolonizing female New Zealand (NZ) sea lions Phocarctos hookeri (n = 13) at Otago Peninsula, South Island, NZ. Human interactions at sea, where sea lions forage, are of concern, and spatial management measures have been proposed. Understanding the level of foraging site fidelity of these animals was consequently essential. We used satellite tracking of individuals across three autumns to assess foraging site fidelity and year-round on-land sighting surveys over 2.5 years as proxy to foraging areas outside autumns. Each individual exhibited a high level of autumnal site fidelity for foraging areas between years (64% overlap between 65% Kernel ranges with a 3-km buffer) while using beaches along a 12-km stretch of coastline during 96 ± 8% (range 79–100%) of their time onshore. As a proxy for foraging areas outside autumns, these animals exhibited a high level of site fidelity to this stretch of coastline throughout the year. Breeding females were sighted there during 86% of months (range = 73–100%) and non-breeding females during 69% of months (range = 58–90%). The site fidelity of these animals indicates that protected areas would be efficient in this case and highlights the importance of studying foraging site fidelity in mobile predators to design efficient conservation measures.