Functional in vivo strain data are examined in relation to bone material properties in an attempt to evaluate the relative importance of osteoporotic bone loss versus fatigue damage accumulation as factors underlying clinical bone fragility. Specifically, does the skeleton have a sufficiently large safety factor (ratio of bone failure strain to maximum functional strain) to require that fatigue damage accumulation is the main factor contributing to increased risk of fracture in the elderly? Existing methods limit in vivo strain measurements to the surfaces of cortical bone. Peak principal compressive strains measured at cortical sites during strenuous activity in various mammalian and avian species range from -1700 to -5200 mu epsilon, averaging -2500 mu epsilon (-0.0025 strain). Much of this threefold variation reflects differences in the intensity of physical activity, as well as differences among species and bones that have been studied. Peak strains can also vary as much as tenfold at different cortical sites within the same bone. No data exist for cortical bone strain during strenuous activity in humans, but it is likely that human bones experience a similar range of peak strain levels. Compact bone fails in longitudinal compression at strains as high as -14,000 to -21,000 mu epsilon, but begins to yield at strains between -6000 and -8000 mu epsilon. Given that yielding involves rapid accumulation of microdamage within the bone, it seems prudent to base skeletal safety factors on the yield strain, rather than the ultimate failure strain of bone tissue. Safety factors to yield failure therefore range from 1.4 to 4.1.(ABSTRACT TRUNCATED AT 250 WORDS)