The irreversible loss of function after axonal injury in the central nervous system (CNS) is a result of the lack of neurogenesis, poor regeneration, and the spread of damage caused by toxicity emanating from the degenerating axons to uninjured neurons in the vicinity. Now, 100 years after Ramon y Cajal's discovery that CNS neurons--unlike neurons of the peripheral nervous system--fail to regenerate, it has become evident that (a) CNS tissue is indeed capable of regenerating, at.least in part, provided that it acquires the appropriate conditions for growth support, and (b) that the spread of damage can be stopped and the postinjury rescue of neurons thus achieved, if ways are found to neutralize the mediators of toxicity, either by inhibiting their action or by increasing tissue resistance to them. In most physiological systems the processes of tissue maintenance and repair depend on the active assistance of immune cells. In the CNS, however, communication with the immune system is restricted. The accumulated evidence from our previous studies suggests that the poor posttraumatic repair and maintenance in the CNS is due at least in part to this restriction. Key factors in the recovery of injured tissues, but missing or deficient in the CNS, are the processes of recruitment and activation of immune cells. We therefore propose the development of immune cell therapies in which the injured CNS is exogenously provided with an adequate number of appropriately activated immune cells (macrophages for regrowth and autoimmune T cells for maintenance), controlled in such a way as to derive maximal benefit with minimal risk of disease. It is expected that these self-adjusting cells will communicate with the damaged tissue, monitor tissue needs, and control the dynamic course of CNS healing.