Malaria-infected red blood cells are under a substantial oxidative stress. Glutathione metabolism may play an important role in antioxidant defense in these cells, as it does in other eukaryotes. In this work, we have determined the levels of reduced and oxidized glutathione (GSH and GSSG, respectively) and their distributions in the parasite, and in the host-cell compartments of human erythrocytes infected with the malaria parasite Plasmodium falciparum. In intact trophozoite-infected erythrocytes, [GSH] is low and [GSSG] is high, compared with the levels in normal erythrocytes. Normal erythrocytes and the parasite compartment display high GSH/GSSG ratios of 321.6 and 284.5, respectively, indicating adequate antioxidant defense. This ratio drops to 26.7 in the host-cell compartment, indicating a forceful oxidant challenge, the low ratios resulting from an increase in GSSG and a decline in GSH concentrations. On the other hand, the concentrations of GSH and GSSG in the parasite compartment remain physiological and comparable to their concentrations in normal red blood cells. This results from de novo glutathione synthesis and its recycling, assisted by the intensive activity of the hexose monophosphate shunt in the parasite. A large efflux of GSSG from infected cells has been observed, its rate being similar from free parasites and from intact infected cells. This result suggests that de novo synthesis by the parasite is the dominating process in infected cells. GSSG efflux from the intact infected cell is more than 60-fold higher than the rate observed in normal erythrocytes, and is mediated by permeability pathways that the parasite induces in the erythrocyte's membrane. The main route for GSSG efflux through the cytoplasmic membrane of the parasite seems to be due to a specific transport system and occurs against a concentration gradient. Gamma-glutamylcysteine [Glu(-Cys)] and GSH can penetrate through the pathways from the extracellular space into the host cytosol, but not into that of the parasite. This implies that the parasite membrane is impermeable to these peptides, and that the host cannot supply GSH to the parasite as suggested previously. Exogenous Glu(-Cys) is not converted into GSH in the host cell, arguing that GSH synthetase may not be functional. Compartment analysis of Mg2+ in infected erythrocytes revealed that the host compartment exhibits a low concentration of Mg2+ (0.5 mM) in comparison with the parasite compartment (4 mM) and the normal erythrocytes (1.5-3 mM). The drop in [Mg2+] results in cessation of Glu(-Cys) synthesis, and hence of GSH synthesis in the host-cell compartment. The decrease in [Mg2+] can affect other Mg2+-ATP-dependent functions, such as Na+ and Ca2+ active efflux. The present investigation confirms that the host-cell compartment is oxidatively distressed, whereas the parasite is efficiently equipped with anti-oxidant means that protect the parasite from the oxidative injury. The parasite has a huge capacity for de novo synthesis of GSH and for the reduction of GSSG. Part of the GSSG that is actively extruded from the parasite is reduced to GSH in the host cell whose own GSH synthesis is crippled.