A series of short- and long-lasting experimental protocols of different meal timing regimes were performed in obese subjects to assess the possible occurrence of (1) a different metabolic fate of nutrients; (2) a phase shift of circadian rhythms of metabolic and hormonal parameters strictly related to nutrition; (3) a different weight loss. (A) In a short-lasting protocol (3 days) 15 obese subjects were fed a hypocaloric diet (684 kcal/day) (a) at 10 hr only, (b) at 1800 hr only; (c) at 1000 hr, 1400 hr and 1800 hr, or (d) studied during a 36-hr fasting. Measures of calorimetry (R.Q., CHO and lipid oxidations, energy expenditure), hormones (plasma cortisol, insulin, HGH, urinary catecholamines), urinary electrolytes (Na, K) and vital parameters (body temperature, heart rate, blood pressure) were carried out at 4-hr intervals for three days. A significantly higher lipid oxidation and a lower CHO oxidation were documented with the meal at 1800 hr, in comparison with the meal at 1000 hr. CHO and lipid oxidation circadian rhythms appeared the most affected by meal timing. (B) In a long-lasting protocol (18 days) 10 obese subjects were fed the same hypocaloric diet (a) at 1000 hr only and (b) at 1800 hr only. Calorimetric measures were performed every other day for 2 hr preceding each meal. Before and after the 18-days single meal period, body temperature, plasma cortisol, PRL and TSH were recorded (delta t = 4 hr). A higher lipid oxidation and a lower CHO oxidation were again demonstrated with the meal at 18 hr. Minimal changes of hormonal circadian rhythms were documented suggesting that the hypothalamus-hypophysis network is scarcely affected by meal timing. Weight loss did not vary in both short- and long-term protocol.