A specific and sensitive double-antibody radioimmunoassay for melatonin (N-acetyl-5-methoxytryptamine) has been developed utilizing rabbit antisera to a bovine serum albumin conjugate of N-succinyl-5-methoxytryptamine and utilizing N-3-(4-hydroxyphenl)-propionyl-5-methoxytryptamine for radioiodination. The least detectable concentration of melatonin standard was 10 pmolar (2.3 pg/tube) with 50% inhibition resultinhibition curves obtained with increasing quantities of melatonin or increasing quantities of chloroform extracts of ovine sera were parallel. The immunoreactivity found in ovine sera c-migrated with [3H]melatonin on silica gel G when developed with chloroform:methanol (9:1). N-Acetylserotonin, 5-methoxytryptamine, serotonin, tryptophan, 6-hydroxymelatonin, 6-methoxytetrahydroharmalan, and several other indole and beta-carboline compounds do not influence the estimation of melatonin in the radioimmunoassay. Concentrations of melatonin could be accurately determined when 31 to 1000 pg were added to 1 ml ovine serum. Serum samples with melatonin concentrations of 1000 pg/ml, 500 pg/ml and 75 pg/ml had intra-assay coefficients of variation of 9.1%, 8.6%, and 17.4%, respectively. The respective inter-assay coefficients of variation were 22.7%, 18.1%, and 37.1%. Ewes exposed to a 12 h light: 12 h dark lighting regimen demonstrated a circadian rhythm in serum concentrations of melatonin. Concentrations ranged from 10-30 pg/ml during periods of light to 100-300 pg/ml during periods of dark. During exposure to continuous light, the circadian rhythm was abolished and concentrations of melatonin were maintained at 10-50 pg/ml. When exposed to conditions of continuous dark the circadian rhythm persisted. A precipitous drop in serum concentrations of melatonin resulted when ewes experiencing peak melatonin concentrations were exposed to light. Concentrations returned to peak levels when the lights were turned off 3.5 h later.