A combination of gene transfer and intracerebral transplantation techniques has been used in studies of CNS development to provide the most compelling evidence to date that the broad diversity of cell types that exist in the CNS arises from single precursor cells. Although the factors that influence cellular differentiation in vivo remain to be clarified, work conducted in vitro with neural precursors has demonstrated that environmental signals (both soluble factors and substrate molecules) play a pivotal role in these decisions. In particular, FGF-2 appears to be one of the prominent influential factors involved in CNS development (see Temple & Qian, 1995). The generation of immortalized precursor populations that are capable of differentiating into multiple CNS cell types in vivo has significant implications for the treatment of neural dysfunction. Such cells may be manipulated toward a lineage that synthesizes factors of interest and used in grafting strategies to replace substances that are lost after injury or in neurodegenerative disease. Alternatively, precursor cells may be directed to a neuronal lineage and used to functionally repair damaged neural systems. Finally, genetic modification of precursor populations provides a method for introducing therapeutic gene products both into discrete regions of the brain and into widely dispersed areas of the CNS. In considering applications to human disease, it has been reported that nestin is expressed in human neuroepithelial cells (Tohyama et al., 1992), suggesting the existence of neural precursors. Recently, such precursors were in fact isolated by two separate groups (Kirschenbaum et al., 1994; Sabaté et al., 1995) and shown to be amenable to gene transfer and to successfully survive transplantation into the brain of experimental animals (Sabaté et al., 1995). Such findings encourage the possibility that precursor cells from the human CNS may be utilized in cell replacement or gene therapy strategies directed toward human neurodegenerative disorders. While immortalization techniques have been essential for generating large quantities of precursor cells for study and transplantation, the genetic modification of cells may alter vital cellular properties. Thus, the ability to induce the proliferation of nonimmortalized neural populations in vitro with the use of growth factors (see section on CNS precursor cells above) provides an important alternative approach for developing perpetual neural cell lines. Recent work with such growth factor-responsive precursor cells has suggested their therapeutic potential in the CNS, as evidenced by the finding that FGF-2-responsive cells can successfully engraft and express transgenes in the adult brain (Gage et al., 1995; Sabaté et al., 1995; Suhonen et al., 1996). Continuing studies with these cells will provide additional insight into the properties of primary CNS stem cells and increase the range of precursor populations that are useful for exploring the development, function, and plasticity of the CNS.