Bone mass can be viewed as the net product of two counteracting metabolic processes, bone formation and bone resorption, which allow the skeleton to carry out its principal functions: mechanical support of the body, calcium dynamic deposition and haemopoiesis. Besides radiological methods, several blood and urinary molecules have been identified as markers of bone metabolic activity for estimating the rates and direction of the biological activities governing bone turnover. The advantages for the use of bone metabolism markers are that they are potentially less dangerous than radiological determinations, are more sensitive to changes in bone metabolism than radiological methods and are easily collected and analysed. The disadvantages are that they have high biological variability. Physical exercise is a known source of bone turnover and is recommended for preventing osteoporosis and bone metabolism problems. There are numerous experiments on bone metabolism markers after acute exercise, but not after long-term training and during or after a whole competition season. Moreover, few studies on bone metabolism markers have evaluated their performance in elite and top-level athletes, who have a higher bone turnover than sedentary individuals. Despite discrepant results among studies, most have shown that short exercise is insufficient for modifying serum concentrations of bone metabolism markers. Marker variations are more evident after several hours or days after exercise, bone formation markers are more sensitive than bone resorption markers, and stimulation of osteoblast and/or osteoclast functions is exercise dependent but the response is not immediate. The response depends on the type of exercise; the markers seem to be less sensitive to resistance exercise and the intensity of exercise is not discriminate. Comparisons between trained subjects and untrained controls have demonstrated the influence of exercise on bone turnover. During training, carboxy-terminal collagen cross-links (CTx), a bone resorption marker, was shown to be less sensitive than amino-terminal cross-linking telopeptide of type I collagen (NTx) and urinary pyridinolines, which were sensitive to anaerobic exercise. Whereas, the bone formation markers, bone alkaline phosphatase (BAP) and osteocalcin (OC) changed after 1 month and 2 months of an exercise programme, respectively. After 2 months, while BAP normalized, it was found to be sensitive to aerobic exercise and OC was found to be sensitive to anaerobic exercise. After prolonged training and competition, bone formation markers are found to change in sedentary subjects enrolled in a physical activity programme. Professional athletes show changes in bone formation markers depending on programme intensity, whereas bone resorption appears to stabilize. Crucial for long-term training, are the characteristics of exercise (e.g. weight-bearing, impact).