Invasive predators are responsible for the extinction of numerous island species worldwide. The naïve prey hypothesis suggests that the lack of co-evolutionary history between native prey and introduced predators results in the absence of behavioral responses to avoid predation. The lack of terrestrial mammal predators is a core feature of islands at the southern end of the Americas. Recently, however, the American mink (Neovison vison) established as a novel terrestrial predator, where rodents became a main portion of its diet. Here, we investigated on Navarino Island, Chile, macro- and micro-habitat selection of small rodents using Sherman traps. Additionally, we experimentally tested behavioral responses of small rodents to indirect cues of native raptorial predation risk (vegetation cover) and direct cues of novel mink predation risk (gland odor) using Sherman traps and foraging trays (giving-up density (GUD)). At the macro-habitat level, we detected native rodents of the species Abrothrix xanthorhinus and Oligoryzomys longicaudatus and the exotic Mus musculus. In general, rodents preferred scrubland habitats. At the micro-habitat level, we only captured individuals of A. xanthorhinus. They preferred covered habitats with tall vegetation. GUD increased in opened areas (riskier for raptorial predation) regardless of the presence or not of mink odor. These results suggest that A. xanthorhinus can perceive predation risk by raptors, but not by mink, results that accord with the hypothesis that co-evolutionary history is important for rodents to develop antipredator behavior. Given that these rodents represent an important proportion of mink diet, the low abundances together with the apparent lack of antipredator response raise conservation concerns for the small rodent populations inhabiting the southernmost island ecosystems of the Americas.