The ability to direct expression of genes to astrocytes in mice has been one of the major motivators of transcriptional analyses of the glial fibrillary acidic protein (GFAP) gene. Another has been the possibility of discovering signaling pathways that operate during development, disease, and injury-all states that increase GFAP gene activity-by identifying and working back from the responsible DNA elements. Here we review studies in both these areas and provide practical guidelines for the construction and analysis of GFAP transgenes. Analyses of the GFAP promoter from cell transfection experiments are summarized to provide background information for the studies in transgenics. Another section provides practical information on the construction and analysis of transgenic mice, with particular reference to GFAP transgenes. The survey of analyses of GFAP transcription elements in transgenic mice reveals that a segment of about 2 kb of the 5'-flanking region of the gene is sufficient to direct reporter gene activity to astrocytes with high specificity. This segment also supports a response to brain injury by upregulation of the activity. Developmentally, the transgene activity is seen by e12.5, several days earlier than GFAP protein or mRNA has been detected. GFAP transcription control regions have already been used to express several proteins in astrocytes to evaluate their biological effects. These proteins include IL-3, IL-6, TGF-beta1, the HIV envelop protein gp120, the MHC Class I Db protein, somatosatin, CNTF, and the herpes simplex virus thymidine kinase. In the future many other GFAP transgenes are expected to be produced, with increasing knowledge of the GFAP regulatory elements promising greater sophistication through promoters that can be regulated, have higher activity, and target activity to particular brain regions.