The focus of attention in osteoporosis research has been on factors influencing bone fragility and the propensity for trauma (falls) during adulthood and old age. The purpose of this review is to change the focus of this attention toward consideration of skeletal growth in the first 20 years of life and to demonstrate the importance of mineral accrual (and the factors influencing this accrual) in determining bone density in adulthood and old age. We suggest that the epidemiology of fractures may be unified by a central role of reduced peak bone density in the pathogenesis of the low bone density found in patients with fractures. The reduced peak bone density establishes the relevance of age-related and sex hormone-dependent bone loss. Risk and protective factors in the first 20 years of life may have quantitatively larger, and qualitatively different effects on the axial and appendicular skeleton than exposure during adulthood. Public health measures focussed on optimising mineral accrual in the first 20 years of life may be more important than health care measures more proximate to the age at which fractures occur.