Herpes simplex virus type 1 represents an ideal candidate for development as a vehicle for gene transfer to postmitotic neurons of the central nervous system. The natural biology of this virus makes it well suited for this purpose as it is capable of infecting a variety of neuronal cell types in the brain where the viral genome can persist indefinitely in a latent state. In latency, the viral lytic genes are transcriptionally silent and a unique set of latency-associated transcripts are expressed. Two impediments to using herpes simplex virus vectors must be overcome: (1) A noncytotoxic mutant virus backbone must be engineered, and (2) a suitable promoter-regulator that stably expresses foreign genes from the vector genome during latency must be constructed. Deletion of specific immediate early genes from the vector can render the virus nontoxic to neurons in culture and in vivo following stereotactic inoculation into specific regions of the brain. Because these viruses cannot replicate, they enter latency on infection of central nervous system neurons. A number of viral and cellular promoters have been tested for their ability to express genes during latency. Strong viral promoters and neurospecific promoters display transient activity. Although the promoter regions for the latency-associated transcripts are highly active in the peripheral nervous system, they show low-level but persistent activity in the brain. Experiments are in progress to exploit RNA polymerase III gene promoters or novel recombinant promoters capable of auto-inducing their own expression in order to increase gene expression during latency in brain neurons.