Abstract: Antioxidant activity of melatonin in human erythrocytes, exposed to oxidative stress by cumene hydroperoxide (cumOOH), was investigated. CumOOH at 300 μM progressively oxidized a 1% suspension of red blood cells (RBCs), leading to 100% hemolysis in 180 min. Malondialdehyde and protein carbonyls in the membrane showed a progressive increase, as a result of the oxidative damage to membrane lipids and proteins, reaching peak values after 30 and 40 min, respectively. The membrane antioxidant vitamin E and the cytosolic reduced glutathione (GSH) were totally depleted in 20 min. As a consequence of the irreversible oxidative damage to hemoglobin (Hb), hemin accumulated into the RBC membrane during 40 min. Sodium dodecyl sulfate (SDS) gel electrophoresis of membrane proteins showed a progressive loss of the cytoskeleton proteins and formation of low molecular weight bands and protein aggregates, with an increment of the intensity of the Hb band. Melatonin at 50 μM strongly enhanced the RBC resistance to oxidative lysis, leading to a 100% hemolysis in 330 min. Melatonin had no effect on the membrane lipid peroxidation, nor prevented the consumption of glutathione (GSH) or vitamin E. However, it completely inhibited the formation of membrane protein carbonyls for 20 min and hemin precipitation for 10 min. The electrophoretic pattern provided further evidence that melatonin delayed modifications to the membrane proteins and to Hb. In addition, RBCs incubated for 15 min with 300 μM cumOOH in the presence of 50 μM melatonin were less susceptible, when submitted to osmotic lysis, than cells incubated in its absence. Extraction and high-performance liquid chromatography (HPLC) analysis showed a much more rapid consumption of melatonin during the first 10 min of incubation, then melatonin slowly decreased up to 30 min and remained stable thereafter. Equilibrium partition experiments showed that 15% of the melatonin in the incubation mixture was recovered in the RBC cytosol, and no melatonin was extracted from RBC membrane. However, 35% of the added melatonin was consumed during RBC oxidation. Hydroxyl radical trapping agents, such as dimethylsulfoxide or mannitol, added into the assay in a 1,000 times molar excess, did not vary melatonin consumption, suggesting that hydroxyl radicals were not involved in the indole consumption. Our results indicate that melatonin is actively taken up into crythrocytes under oxidative stress, and is consumed in the defence of the cell, delaying Hb denaturation and release of hemin. RBCs are highly exposed to oxygen and can be a site for radical formation, under pathological conditions, which results in their destruction. A protective role of melatonin should be explored in hemolytic diseases.