The beta-thalassemias are congenital anemias that are caused by mutations that reduce or abolish expression of the beta-globin gene. They can be cured by allogeneic hematopoietic stem cell (HSC) transplantation, but this therapeutic option is not available to most patients. The transfer of a regulated beta-globin gene in autologous HSCs is a highly attractive alternative treatment. This strategy, which is simple in principle, raises major challenges in terms of controlling expression of the globin transgene, which ideally should be erythroid specific, differentiation- and stage-restricted, elevated, position independent, and sustained over time. Using lentiviral vectors, May et al. demonstrated in 2000 that an optimized combination of proximal and distal transcriptional control elements permits lineage-specific and elevated beta-globin expression, resulting in therapeutic hemoglobin production and correction of anemia in beta-thalassemic mice. Several groups have by now replicated and extended these findings to various mouse models of severe hemoglobinopathies, thus fueling enthusiasm for a potential treatment of beta-thalassemia based on globin gene transfer. Current investigation focuses on safety issues and the need for improved vector production methodologies. The safe implementation of stem cell-based gene therapy requires the prevention of the formation of replication-competent viral genomes and minimization of the risk of insertional oncogenesis. Importantly, globin vectors, in which transcriptional activity is highly restricted, have a lesser risk of activating oncogenes in hematopoietic progenitors than non-tissue-specific vectors, by virtue of their late-stage erythroid specificity. As such, they provide a general paradigm for improving vector safety in stem cell-based gene therapy.