An ecosystem engineer is an organism that creates or modifies habitats for other species. Numerous research has reported that the abundance of species in the physical structures created by ecosystem engineers—referred to as engineered habitats—is higher than in other typical habitats—referred to as non-engineered habitats—suggesting the importance of the roles ecosystem engineers play in the population demography of species utilizing engineered habitats (i.e., utilizing species). However, it remains largely unknown whether engineered habitats, in comparison to non-engineered habitats, are more suitable for the survival and growth of the utilizing species. In this study, we addressed this issue by assessing the suitability of wallow, created by Japanese sika deer (Cervus nippon yesoensis), as an alternative larval habitat for Rana pirica frogs and Hynobius retardatus salamanders that oviposit their eggs in the deer wallows. We compared the early life history performance (i.e., from egg to mid-larval stage) of the two amphibian species in the deer wallows and in non-engineered natural ponds. Our survey revealed that R. pirica frogs survived better in wallows than in ponds, while H. retardatus salamanders survived better in ponds than in wallows. Additionally, R. pirica frogs had a smaller size at metamorphosis and a longer time until metamorphosis in wallows than in ponds. The species-specificity (i.e., survival of frogs vs. survival of salamanders) and the trait-specificity (i.e., survival vs. growth of frogs) in the suitability of deer wallows represent the complexity of the effects of ecosystem engineers on the utilizing species. For example, the contrasting suitability of deer wallows for the survival of the two amphibian species suggests that sika deer may have distinctive effects on the population demography of the two species by creating wallows that function as refuges and ecological traps for R. pirica frogs and H. retardatus salamanders, respectively.