No prospective randomized study with fracture as end point exists in men. Data from retrospective and prospective observational and case control studies suggest that activity in men is associated with reduced fracture risk. This may be correct, but consistently replicated sampling bias may produce the same observation, as these studies are hypothesis generating, never hypothesis testing. Higher musculoskeletal mass, better health, coordination, and less tendency to fall may lead to exercise, not the reverse. It would be extremely difficult to conduct an exercise intervention study with fracture as the end point because of the large cohorts needed. However, showing a positive effect on surrogate end points for fractures as increased bone mass or reduced fall frequency would support the notion that exercise has a fracture-protective effect. Exercise during growth seems to build a larger and stronger skeleton in boys. However, cessation of exercise is its Achilles heel; biologically important increased peak bone mass or improvement in muscle strength achieved by exercise during growth may be eroded in retirement, leaving no biological significant benefits in old age when fractures occur. Exercise during adulthood may prevent bone loss or produce a small increase in BMD of a few percent, probably a non-biologically significant increase in reducing the fracture risk in elderly men. However, exercise in adults seems to increase muscle strength and improve balance, with reduced fall frequency as the result and maybe reduced frequency of injuries also. The effects of exercise on bone size, shape, and architecture during growth and adolescence must be defined as well as the effect of muscle function, fall frequency, and frequency of injurious falls in elderly men. Also, continued low level of exercise may maintain some of the benefit after more vigorous activity level during younger years, but dose-response relationships need to quantified. Additionally, the null hypothesis that exercise has no effect on fracture rates in old age cannot be rejected by any published data. The proof rests on demonstration of a reduction in spine and hip fractures in well-designed and executed prospective, open-randomized studies, none of which exist. Our inability to answer these questions should be acknowledged before recommendations are made at the community level.