The fundamental message emerging from immunologic and immunopathologic analyses of the brain and spinal cord in multiple sclerosis (MS) is that during inflammation, the central nervous system (CNS) is capable of interactions with the lymphoid system, mainly through induced (as opposed to constitutive) expression of immune system-specific molecules on CNS elements. CNS endothelium, astrocytes and microglial cells are the main participants, with oligodendrocytes and neurons remaining essentially inert. There appears to be nothing unique about the manner in which the CNS responds to inflammation or in the molecules expressed. The ensuing adhesion molecules, pro-inflammatory and regulatory cytokines, histocompatibility molecules, and T and B cell markers, are difficult to distinguish from those occurring in peripheral lymphoid tissue. However, differences certainly exist in the outcome of an inflammatory insult in the CNS versus other, peripheral tissues, whereby there is generally a poor reparatory response. Reasons for the latter appear to lie in the anatomical complexity of the CNS, its vulnerability to damage by soluble mediators, and in the white matter (the battlefield for the inflammatory attack in MS), the exquisite sensitivity of the oligodendrocyte and its myelin to exogenous factors. With the aid of examples drawn from experimental allergic encephalomyelitis, the prime animal model for MS, a number of approaches to prevent or downregulate CNS inflammation during immune-mediated demyelination are presented as possible therapeutic avenues for MS, some of which are already under investigation.