The continuing ability of the skeleton to withstand functional loads without damage requires that bone mass and architecture are adjusted according to the loads experienced. Load bearing is the only functional influence that requires a particular bone architecture, and functionally engendered strains within the bone tissue provide the only feedback containing the necessary information on the relationship between current architecture and prevailing load history. The specific strain-related objectives of the adaptive modeling and remodeling response to load bearing have not been adequately defined. They appear to be different for cortical and cancellous bone and vary according to cortical location. Experiments suggest that adaptive modeling and remodeling is sensitive to dynamic but not static strain change and that the osteogenic response to a period of dynamic strain is quickly saturated but is higher when the rate of change in strain is high and the distribution of strain unusual. Presumably it is the cumulative effect of this osteogenic response to load bearing that normally maintains bone mass above that seen in disuse situations. Through their independent effects on bone cell behavior, nutritional and hormonal factors can enable, enhance, limit, or frustrate full expression of the osteogenic response to strain change. However, such systemic factors do not appear to be able to engender or successfully imitate the sustained cumulative local response to load bearing that normally maintains functionally appropriate bone mass and architecture. Experiments in vivo and in vitro suggest that in osteocytes and surface osteoblasts the almost immediate response to strain change is increased production of prostacyclin. Surface osteoblasts also produce prostaglandin E. Only 5 minutes after loading, glucoses-phosphate dehydrogenase activity in osteocytes is increased in a local strain magnitude-related manner, and 24 h later there is an increase in osteocyte RNA. Exogenous PGE2 and PGI2 imitate the G6PD response in both osteocytes and osteoblasts, but only PGI2 imitates the loading-related increase in RNA in these two cell types. Indomethacin reduces the osteogenic response to loading in vivo and both the G6PD and RNA responses to loading in vitro. We hypothesize that the strain-related increase in PGE influences the synthetic activity of surface bone cells directly, whereas the strain-related increase in PGI2 additionally influences modeling and remodeling through the production of a cytokine or growth factor for which the loading-related RNA is coded. It may be at the stage of cytokine interaction that the potentially competing or complementary effects on modeling and remodeling of the loading and hormonal environments is resolved.