We have examined aspects of methemoglobin (metHb) reduction in sickle and in thalassemic red blood cells (RBCs). NADH metHb reductase activity in sickle and thalassemic RBCs was significantly increased compared with normal RBCs. Because in vitro enzyme activity does not necessarily represent in vivo activity, we measured the rate of metHb reduction in intact RBCs. Intact thalassemic RBCs demonstrated a significantly increased rate of metHb reduction compared with normal RBCs. In contrast, intact sickle RBCs had a rate of metHb reduction that was similar to normal RBCs and significantly decreased relative to high reticulocyte RBCs of equivalent cell age. To determine the mechanism for the relative impairment of metHb reduction in sickle RBCs, we measured intraerythrocytic NADH, a cofactor in the metHb reduction reaction. Thalassemic RBCs had a significantly increased NADH content relative to normal RBCs. In contrast, sickle RBCs did not have an increase in NADH content. Furthermore, incubating normal RBCs under conditions that increase the NADH content resulted in an increased rate of metHb reduction. In contrast, conditions that decrease the NADH content in normal RBC resulted in a decreased rate of metHb reduction. These data and other results suggest that metHb reduction in intact RBCs is dependent on NADH content, and that the impaired metHb reduction rate in sickle RBCs may be a result of a lack of increase in NADH content. The dependence of metHb reduction on RBC NADH content and the ability to manipulate NADH content in vitro suggest a new strategy for decreasing oxidant damage to sickle RBCs in vivo.