Melatonin is natural occurring hormone biosynthesized from the pineal gland of vertebrate. It is also identified in various kinds of plants. There are several potential mechanisms proposed for the melatonin including antioxidative effect. Results from previous studies had demonstrated an increasing of serum melatonin concentration following fruit consumption. Nonetheless, the correlation between the amount of tropical fruit consumption and serum melatonin concentration is currently limited. The crossover study was intentionally conducted in healthy volunteers to determine of serum concentration of melatonin. Participants were assigned to take one of the following fruit preparations, juice extracted from one kilogram of orange or that of pineapple or 2 bananas, with one week wash-out period for each kind of fruit consumption. Enzyme-linked immunosorbent (ELISA) assay was used for measuring the serum melatonin concentration. Antioxidant capacity was determined by Ferric reducing antioxidant power (FRAP) assay and Oxygen radical antioxidant capacity (ORAC) assay. Twelve young males were enrolled in the study. Serum melatonin concentration at 120 minutes after fruit consumption compared with before consumption were significantly increased in pineapples (145.31 vs 47.95 pg/mL p=0.002), orange (151.02 vs 39.94 pg/mL, p=0.005) and banana (140.33 vs 32.37 pg/mL, p=0.008), respectively. Corresponding to serum melatonin concentration, the serum antioxidant capacity following fruit consumption was also significantly increased in both assays. For FRAP assay the antioxidant capacity in pineapple (12.61% increase, p=0.003), orange (7.72% increase, p=0.004) and banana (14.26% increase, p=0.002), respectively. For ORAC assay the antioxidant capacity increase in pineapple (8.62% increase, p=0.002), orange (5.74% increase, p=0.002) and banana (8.35% increase, p=0.002), respectively. These findings suggest that tropical fruit consumption can increase the serum melatonin concentration and could potentially demonstrate the antioxidant effect in healthy volunteers