Captive specimens in natural history collections allow researchers to inspect the morphologies of rare taxa, but the lifestyles, diets, and lifespans of captive animals differ from those of their wild counterparts. To quantify these differences, we compared bone microstructure of trunk vertebrae in captive and wild xenarthran mammals (sloths, armadillos, anteaters). Because trabecular bone architecture (TBA) adapts to in vivo forces, bone microstructure reflects ecology and behavior, but this means that it may differ between captive and wild specimens of the same species. We collected μCT scans of the last six presacral vertebrae in 13 species of fossorial, terrestrial, and suspensorial xenarthrans ranging in body mass from 120g (Chlamyphorus) to 35kg (Myrmecophaga). For each vertebra, we measured bone volume fraction (BVF); trabecular number, mean thickness (TbTh), and orientation; global compactness; and cross sectional area. Wild specimens generally have more robust trabeculae, but this differs based on species, vertebral position, ecology, and pathology. The wild specimens of fossorial taxa (Dasypus) have more robust trabeculae than their captive counterparts, but there is no clear difference in TBA of wild and captive specimens in suspensorial and terrestrial taxa (Bradypus, Choloepus, Cyclopes). These data suggest that locomotor ecology affects the level to which captivity affects bone microstructure. The captive specimens of both Tamandua and Myrmecophaga have higher BVF and TbTh than their wild counterparts, indicating more brittle trabeculae due to bone pathologies caused by captivity. Our results add to the overall understanding of variation in mammalian bone microstructure and suggest caution when including captive specimens in research on TBA.