The efficacy of osteoporosis therapy can be considered at several different levels, ranging from the overall community to the individual patient. Efficacy has different connotations for each group within the community and for the individual, in whom life expectancy--among other parameters--is an important consideration. Some of the questions addressed in this article are the scope of the osteoporosis problem, how many people must be treated for prevention of one further fracture, whether there is a role for targeting likely high risk/high response groups, and what exactly is a successful outcome. Clearly, there are no ideal therapies as yet for osteoporosis, in the context of either prevention or treatment. Moreover, most studies have been performed in selected groups within the community, and the extrapolation to wider clinical subsets is not straightforward. For example, data are almost nonexistent for men or premenopausal women; also it is not clear whether studies on carefully selected healthy elderly women can be simply extrapolated to their less healthy cohorts. Similarly, studies using vertebral deformities as the primary endpoint may be difficult to relate to symptomatic vertebral fractures, and other peripheral fractures including proximal femur. Lastly, the efficacy of some treatments must be related to the risk of further fractures. Thus, many years of therapy may relate to prevention of perhaps only one symptomatic event, with equivocal cost-benefit for the individual. As new therapies are under development, these questions need to be addressed for the efficacy of osteoporosis therapy to be considered.