Serum osteocalcin (OC), which is a sensitive marker of bone formation, is reduced during chronic glucocorticoid treatment in accordance with the finding of reduced bone formation, and even short term glucocorticoid treatment reduces serum OC. In a double blind placebo-controlled study, we measured the effects of 2.5 and 10 mg prednisone, orally, on the circadian variation of serum OC in 15 normal subjects (8 women and 7 men), aged 22-46 yr. All subjects were studied twice at an interval of 1 week. Blood samples were collected every 60 min from 1630 until 1700 h the following day. Prednisone or placebo was given at 2000 h. Serum OC was measured by an in-house RIA. After placebo administration serum OC increased from 2230 h to a peak value around 0230 h, followed by a gradual decline to a nadir around 1500 h. Both doses of prednisone inhibited and even reversed the nocturnal rise in serum OC levels; the inhibition occurred within 3-4 h. There was no significant difference in the time from prednisone ingestion to maximal decrease (7.2 h vs. 9.8 h) or in the maximal decrease (52% vs. 54%). However, the duration of the inhibition was significantly shorter after 2.5 mg than after 10 mg prednisone (approximately 6 h vs. 12 h; (p less than 0.01). We conclude that serum OC is sensitive to small doses of prednisone, and that serial serum OC measurements may be a sensitive marker of the effect of exogenous glucocorticoids on osteoblastic activity in vivo.