We previously demonstrated that daily melatonin administration to middle-aged rats to restore youthful plasma melatonin levels also decreased body weight, visceral fat, plasma leptin, and plasma insulin to more youthful levels, without detectable changes in consumption of chow diet. We now evaluate: (a) whether melatonin alters consumption of a more precisely quantifiable liquid diet similar in high-fat content to the typical American diet; (b) differences between melatonin-induced endocrine responses in the fasted vs fed state; and (c) time course of these responses. Ten-month-old male Sprague- Dawley rats received liquid diet containing either 0.2 micro g/mL melatonin (MELATONIN) or vehicle (CONTROL) (n = 14/treatment); the diet was available throughout each night, but was removed for the final 10 h of each daytime. MELATONIN rats gained 4% body weight during the first 2 wk and then stabilized, whereas CONTROL rats continued to gain for an additional week, achieving 8% gain (p < 0.05 vs MELATONIN). During the first 3 wk, afternoon tail-blood leptin, but not insulin, levels decreased in melatonin-treated rats (p < 0.05 vs CONTROL). After 8 wk, half of the rats were killed at the midpoint of the dark period (NIGHT; fed) and half at the end of the light period (DAYTIME; fasted). NIGHT but not DAYTIME plasma leptin levels were decreased in MELATONIN rats, whereas DAYTIME but not NIGHT plasma insulin levels were decreased (p < 0.05 vs CONTROL). Melatonin treatment did not alter cumulative food consumption. Thus, melatonin decreased weight gain in response to high-fat diet, decreased plasma leptin levels within 3 wk-before decreasing plasma insulin-and exerted these metabolic effects independent of total food consumption.