To evaluate the effect of prednisone and triple treatment (sodium fluoride, calcium, and vitamin D) on trabecular and cortical bone serial bone mineral content (BMC) measurements were made at a metaphyseal (BMCD) and diaphyseal (BMCP) site on the forearm on 31 consecutive and previously bone-healthy patients scheduled for at least 24 weeks high-dose prednisone treatment. The patients were randomized into two further treatment groups: group I (n = 16) received prednisone plus triple treatment and group II (n = 15) received only prednisone. The two groups were similar with regard to age, sex, prednisone dose, and initial BMC. During 24 weeks treatment, BMCD (partially representing trabecular bone) and BMCP (mainly representing cortical bone) fell significantly and similarly, demonstrating that there is no preventive effect on bone mineral loss on the triple regimen. The BMC fall after 12 weeks was significantly more pronounced for metaphyseal (partially trabecular) than for diaphyseal (cortical) bone, whereas the values did not differ significantly after 24 weeks; this indicates a greater sensitivity to the hormone treatment of trabecular bone. In the entire group, the fall in BMC correlated positively with individual prednisone dose, significant at the diaphyseal site (r = 0.39, P less than 0.05), but not at the metaphyseal site (r = 0.31, P = 0.08). It is concluded that corticosteroid-induced osteopenia is a diffuse bone disease which affects trabecular as well as cortical bone, suggesting that BMC measured on the forearm reflects changes in bone mineral at other locations.