Objectives One of the hallmarks of contemporary osteoporosis and bone loss is dramatically higher prevalence of loss and fragility in females post‐menopause. In contrast, bioarchaeological studies of bone loss have found a greater diversity of age‐ and sex‐related patterns of bone loss in past populations. We argue that the differing findings may relate to the fact that most studies use only a single methodology to quantify bone loss and do not account for the heterogeneity and complexity of bone maintenance across the skeleton and over the life course. Methods We test the hypothesis that bone mass and maintenance in trabecular bone sites versus cortical bone sites will show differing patterns of age‐related bone loss, with cortical bone sites showing sex difference in bone loss that are similar to contemporary Western populations, and trabecular bone loss at earlier ages. We investigated this hypothesis in the Imperial Roman population of Velia using three methods: radiogrammetry of the second metacarpal (N = 71), bone histology of ribs (N = 70), and computerized tomography of trabecular bone architecture (N = 47). All three methods were used to explore sex and age differences in patterns of bone loss. Results The suite of methods utilized reveal differences in the timing of bone loss with age, but all methods found no statistically significant differences in age‐related bone loss. Discussion We argue that a multi‐method approach reduces the influence of confounding factors by building a reconstruction of bone turnover over the life cycle that a limited single‐method project cannot provide. The implications of using multiple methods beyond studies of bone loss are also discussed.