Malaria still is a major public health problem, partly because the pathogenesis of its major complication, cerebral malaria, remains incompletely understood. Experimental models represent useful tools to better understand the mechanisms of this syndrome. Here, data generated by several models are reviewed both in vivo and in vitro; we propose that some pathogenic mechanisms, drawn from data obtained from experiments in a mouse model, may be instrumental in humans. In particular, tumor necrosis factor (TNF) receptor 2 is involved in this syndrome, implying that the transmembrane form of TNF may be more important than the soluble form of the cytokine. It has also been shown that in addition to differences in immune responsiveness between genetically resistant and susceptible mice, there are marked differences at the level of the target cell of the lesion, namely, the brain endothelial cell. In murine cerebral malaria, a paradoxical role of platelets has been proposed. Indeed, platelets appear to be pathogenic rather than protective in inflammatory conditions because they can potentiate the deleterious effects of TNF. More recently, it has been shown that interactions among platelets, leukocytes, and endothelial cells have phenotypic and functional consequences for the endothelial cells. A better understanding of these complex interactions leading to vascular injury will help improve the outcome of cerebral malaria