We know little regarding how specific aspects of habitat influence spatial variation in site occupancy by Arctic wildlife, yet this information is fundamental to effective conservation. To address this information gap, we assessed occupancy of 84 Gyrfalcon (Falco rusticolus Linnaeus, 1758) breeding territories observed annually between 2004 and 2013 in western Alaska. In line with the theory of population regulation by site dependence, we asked whether Gyrfalcons exhibited a nonrandom pattern of site selection and if heterogeneous landscape attributes correlated with observed occupancy patterns. We characterized high- and low-occupancy breeding territories as those occupied more or less often than expected by chance, and we evaluated land cover at 1 and 15 km circles centered around nesting territories to identify habitat variables associated with observed occupancy patterns. We tested 15 competing models to rank hypotheses reflecting prey and habitat variables important to nesting Gyrfalcons. We confirmed a nonrandom pattern of site selection but found only weak evidence that the distribution of prey habitat was responsible for this pattern. We reason that preferential habitat use by nesting Gyrfalcons may be determined by spatial scales other than those we measured or may be driven by landscape-level attributes at time periods other than during the brood rearing period.