Evidence in human studies of the association of long term habitual exercise with bone mineral content (BMC) and density (BMD) largely comes from studies in athletes. In young adults, the highest BMC and BMD values have been found in strength and power-trained athletes, while endurance activities such as long distance running and swimming seem less effective with regard to peak bone density. Intensive endurance training may even be associated with amenorrhoea and decreased trabecular bone density in young females. However, after menopause female athletes show greater bone mass indicating that they do not share the accelerated decline in BMC observed in a nonathletic population. Middle-aged and elderly male athletes from various sports have significantly higher BMC and BMD than controls, especially in trabecular bone sites, but higher cortical BMC has also been found in the dominant/nondominant arm comparisons with unilateral exercises such as tennis. The differences found between female athletes and controls have generally been less pronounced than those among men, but a number of studies suggest that in women long term physical training may counteract the low BMC and BMD associated with reduced bone mass. Although the interpretation of results of cross-sectional studies should be treated with caution, studies in athletes serve as an economical alternative approach to experimental trials with their long term follow-up and exercise compliance problems. The differences found in BMD between those who have devoted themselves to life-long training and those who have been much less active should not be underestimated.