This study assessed whether relatives with low bone mineral density (BMD) could be identified in five large families using historical, biochemical, and genetic markers for osteoporosis. Fifty of 65 relatives had their bone density and bone turnover markers measured, together with an assessment of their risk factors for osteoporosis. Only 33% (5/15) of siblings, 50% (6/12) of children and 43% (10/23) of nephews and nieces had entirely normal BMD. There was no difference in life-style risk factors for osteoporosis, history of previous fractures or body mass index between normal subjects and those with osteopenia or osteoporosis. Osteopenic individuals had a significantly higher than normal osteocalcin value. Within families, there was no clear association between BMD and any of the genetic markers (vitamin D receptor gene polymorphisms, COL 1A1 and COL 1A2 polymorphisms of the collagen gene), either alone or in combination. The addition of genetic markers to the other risk factors for low BMD did not improve the prediction of BMD. In conclusion, we suggest that the presence of osteoporosis in a first degree relative should be one of the clinical indications for bone density measurement as the individuals at risk would not be picked up by other methods.