According to certain "reference ranges", about a billion people worldwide, including 80% of the French population, have vitamin D insufficiency or deficiency. In France, serum vitamin D assay has become commonplace. Between 2010 and 2011, the number of tests performed increased by about 70%, the most rapid growth in any laboratory assay. How are "normal" vitamin D levels and, thus, vitamin D insufficiency or deficiency, established, and what is their clinical relevance? To answer these questions, we reviewed the literature using the standard Prescrire methodology. In early 2013, there is no international consensus on normal serum concentrations of 25-hydroxyvitamin D (25(OH)D). Using biochemical criteria, some experts define normal vitamin D status as a blood 25(OH)D concentration above 30 ng/ml and vitamin D insufficiency as levels between 20 and 30 ng/ml. The definition of vitamin D insufficiency is disputed by other experts, who point out that 97.5% of the general population with vitamin D levels of at least 20 ng/ml have no bone disorders. Biostatistical data suggest that the normal range of 25(OH)D levels is between 11 and 46 ng/ml, but the lower limit appears to vary according to the study population. Epidemiological data have shown a link between 25(OH)D concentrations below 20 ng/ml and a variety of clinical disorders, including osteoporosis, fractures, cancer, infections, and autoimmune diseases. Although these data suggest a possible increase in the risk of particular disorders, they provide no information on the harm-benefit balance of vitamin D supplementation. A few clinical trials have shown that combined calcium and vitamin D supplementation (but not vitamin Dalone) has a modest preventive effect on osteoporotic fractures in some subgroups of persons over 65 years old. Some trials showed that the calcium plus vitamin D combination was more effective than calcium alone or placebo in reducing falls among some elderly subgroups. Elderly patients were recruited for most of these trials regardless of their vitamin D status. Vitamin Dsupplementation carries a risk, albeit very small, of vitamin D overdose and hypercalcaemia. Vitamin D plus calcium supplementation increases the risk of kidney stones. In practice, when treatment with vitamin D is envisaged, there is no need to determine vitamin D status by conducting blood tests. In contrast, monitoring calcium levels can help prevent overdose.