Osteoporosis is a common condition, which is recognised by the occurrence of fragility fractures and leads to considerable mortality and morbidity with huge financial implications world-wide. Based on predicted demographic changes, the implications of this disease are set to increasingly affect the healthcare budgets of all nations. The determinants of fracture are skeletal factors, such as peak bone mass, the rate of bone loss and extra-skeletal factors, which include trauma and the response to that trauma. Some of these factors are genetically determined, but several have environmental origins, which could, theoretically, be manipulated. There are two potential means whereby osteoporotic fractures might be prevented. Measures could be targeted at the entire population, with the aim of shifting the distribution of bone mass in a beneficial direction, through modifying the behaviour of all individuals. The alternative is a high risk approach, whereby intervention is targeted only at those considered to have the greatest risk of future fracture. Mass bone density screening falls into the second approach. Bone density is a good predictor of future fracture risk, and cost-effectiveness analyses of the high risk approach suggest economic benefits of policies targeting pharmacological treatment to those individuals at highest risk. However, there are important concerns about the levels of compliance achievable with such therapeutic interventions, the balance of risks and benefits for some of these interventions (for example, hormone replacement therapy), and the outcome when treatment is discontinued. On current evidence, it is certainly not appropriate to target hormone replacement therapy for women at the menopause on the basis of a bone density screening programme. However, newer bone-specific agents are being developed which might be administered at later ages, closer to the time when fracture incidence rates rise steeply. Bone densitometry has been shown to predict fractures even in the elderly, and high risk strategies for the targeting of such agents (for example, the bisphosphonates or selective oestrogen receptor modulators) will remain important research issues for the future.