We studied the acquisition of bone mineral in 45 healthy prepubertal and pubertal girls and related changes in bone mass to age, body mass, pubertal status, calcium intake, and exercise. A subgroup of 12 girls was followed longitudinally. Bone mineral content (BMC) of the lumbar spine, whole body, and femoral neck was measured by dual energy x-ray absorptiometry and that at the midradius by single photon absorptiometry. For comparison, spine and whole body mineral contents were also measured by dual photon absorptiometry. Bone mass was expressed in conventional terms of BMC and area density (BMD). However, we show that BMD fails to account for differences in bone thickness. Since bone size increases during adolescence, we present a new expression, bone mineral apparent density (BMAD), which is BMC normalized to a derived bone reference volume. This term minimizes the effect of bone geometry and allows comparisons of mineral status among bones of similar shape but different size. BMC increased with age at all sites. These increases were most rapid in the early teens and plateaued after 16 yr of age. When bone mineral values at all sites were regresssed against age, height, weight, or pubertal stage, consistent relationships emerged, in which BMC was most strongly correlated, BMD was correlated to an intermediate degree, and BMAD correlated only modestly or without significance. Dietary calcium and exercise level did not correlate significantly with bone mass. From these relationships, we attribute 50% of the pubertal increase in spine mineral and 99% of the change in whole body mineral to bone expansion rather than to an increase in bone mineral per unit volume. In multiple regressions, pubertal stage most consistently predicted mineral status. This study emphasizes the importance of pubertal development and body size as determinants of bone acquisition in girls. BMAD may prove to be particularly useful in studies of bone acquisition during periods of rapid skeletal growth.