Exercise is associated with increased peak bone mineral density (BMD). To determine the relationship between the duration of exercise and BMD, we measured BMD of the axial and appendicular skeleton by dual-energy X-ray absorptiometry (DXA), and speed of sound (SOS), broadband attenuation (BUA), and stiffness index by quantitative ultrasound (QUS) of the calcaneus, in 67 active male national soccer players (mean age 23 years, range 17-35), which included 23 premier-league players exercising 12 h/week (range 8-18), 23 third-league players exercising 8 h/week (range 3-18), and 21 sixth-league players exercising 6 h/week (range 2-10). Results were compared with 24 sedentary age- and gender-matched controls and presented as mean +/- SEM. BMD was higher in all weight-bearing regions for the whole group relative to controls (BMD: total body 6.8 +/- 0.7%, leg 9.6 +/- 0.8%, lumbar spine 13.2 +/- 1.2%, femoral neck 12.7 +/- 1.2% [all p < 0.001]; calcaneus SOS 4.2 +/- 0.3%, BUA 8.7 +/- 1.5%, and stiffness index 24.2 +/- 2.0% [all p < 0.01]). No differences were found in head or arm BMD. There were no differences in BMD or QUS measurements when comparing soccer players exercising for different activity durations. Duration of activity correlated with BMD weight-loaded regions and with QUS, provided it was less <6 h/week (p < 0.01 respectively), but not when exercising more frequently. Femoral neck BMD increased by 3.3% across every hour increase in activity in those with 0-6 h of exercise/week and by 0.7% in those exercising more than this (p < 0.01). We conclude that, in national-league soccer, the BMD needed to attain a bone strength commensurate with that of duration of activity is achieved by 6 h of exercise per week. Beyond this, additional exercise confers no higher BMD. The skeleton adapts to the prevalent level of exercise intensity required and no further.