The relevance of measuring urinary melatonin (MLT) for human pineal research is sometimes questioned, and the relationship among serum levels of MLT, urinary excretion of the unmetabolized hormone, and excretion of MLTs main metabolite, 6-hydroxymelatonin sulfate (aMT6s), is still uncertain. We applied a well established RIA for measuring MLT in serum to urine samples, characterized its criteria of performance in this body fluid, and used it for human studies. In 16 adolescents, the endogenous overnight MLT secretion, expressed as the area under the concentration time curve, correlated significantly with the amounts of urinary aMT6s (r = 0.86; P < 0.0001) and urinary MLT (r = 0.70; P = 0.0027) excreted during a 16-h observation period. Oral administration of 3 mg exogenous MLT in 17 healthy volunteers resulted in peak MLT serum levels differing 28-fold among subjects (940-27,240 pg/ mL; range). In this study urinary MLT, but not aMT6s, excretion was associated with blood MLT concentrations (r = 0.76; P = 0.0004 vs. r = 0.02; P = 0.93, respectively). Thus, endogenous MLT production can be assessed accurately by measuring either aMT6s or MLT excretion. After oral application of MLT, however, only measurement of MLT excretion is a reliable marker of serum concentrations. Determination of MLT in urine may prove to be a useful tool for drug monitoring after oral administration of the pineal hormone.