We administered crystalline melatonin (80 mg) in gelatin capsules to 5 young male volunteers and measured serum and urinary melatonin levels at intervals. Changes in serum melatonin levels were best described by a biexponential equation with an absorption constant (ka) of 1.72 h-1 (half-life = 0.40 h) and an elimination constant (ke1) of 0.87 h-1 (half-life = 0.80 h). Peak serum melatonin levels, ranging from 350 to 10,000 times those occurring physiologically at nighttime, were observed 60-150 min after its administration, remaining stable for approximately 1.5 h. The fraction of ingested melatonin that was absorbed, estimated from the area under the curve describing serum melatonin concentrations as a function of time after melatonin administration (the concentration-time curve), varied by 25-fold among subjects. 3 additional volunteers received three melatonin-containing capsules (80 mg each) at 60-min intervals. This regimen extended the duration of elevated serum melatonin levels to 4-6 h. Melatonin excretion closely paralleled serum melatonin levels until 9 h after the hormone's administration, after which urinary levels tended to be higher than those predicted from serum levels. However, the area under the concentration-time curve for serum melatonin correlated well (r = 0.96) with the cumulative melatonin excretion during the initial 15 h after melatonin's administration, indicating that either approach can be used to estimate the absorption of orally administered melatonin.