Glucocorticoids (GC) induce apoptosis in lymphoblasts and are thus essential in the treatment of acute lymphoblastic leukemia (ALL). Their effects result from gene regulations via the GC receptor (NR3C1/GR), but it is unknown how these changes evolve, what the primary GR targets are, and to what extent responses differ between ALL subtypes and nonlymphoid malignancies. We delineated the transcriptional response to GC on the exon level in a time-resolved manner in a precursor B- and a T childhood ALL model employing Exon microarrays and combined this with genome-wide NR3C1-binding site detection using chromatin immunoprecipitation-on-chip technology. This integrative approach showed that the response was strongly influenced by kinetics and extent of GR autoinduction in both models. Although remarkable differences between the ALL systems were apparent, we defined a set of common response genes enriched in apoptosis-related processes. Globally, GR binding was higher for GC-induced vs. -repressed genes, suggesting that GR mediates gene repression by interaction with distant enhancers or by cross talk with other transcription factors. Exon level analysis defined several new GC-regulated transcript variants of genes, including ATP4B, GPR98, TBCD, and ZBTB16. Our study provides unprecedented insight into the transcriptional response to GC in ALL cells, essential to understand this biologically and clinically important phenomenon. We found evidence of cell type-specific as well as common responses, possibly related to apoptosis induction, and detected induction of novel transcript variants by GC in the investigated systems. Finally, we implemented a bioinformatic framework that might be useful for high-density microarray analyses to identify alternative transcript variant expression.