Despite the availability of guidelines from the World Health Organization study group for the classification of osteoporosis in postmenopausal Caucasian women, confusion still exists about the number of sites used for diagnosis and the clinical utility of peripheral bone mass assessments. To examine the diagnosis of osteoporosis and osteopenia based on bone density measurements at single or multiple sites using central and peripheral measurements, we studied 115 ambulatory, community-dwelling, Caucasian postmenopausal women. Bone mineral density of the hip, PA spine, forearm, and finger were assessed by dual X-ray absorptiometry. Bone mass of the calcaneus was obtained using ultrasound. The diagnosis of osteoporosis based on a single measurement varied from 4% using the trochanteric region to 34% using Ward's triangle, 17% using the calcaneus, and 13% using the finger. Twenty-eight percent of the women had osteoporosis if the diagnosis was based on at least one osteoporotic value at three standard central sites (PA spine, total hip, femoral neck). Among these women, using T-scores provided by the manufacturers, 16% of osteoporotic patients would be misclassified as normal using the Sahara Clinical Bone Sonometer (Hologic, Waltham, MA) (heel) and 34% misclassified using the accuDEXA (Schick, New York, NY) (finger). We conclude that there is significant variability in the classification of osteoporosis based on site selection, with significant potential for misdiagnosis.