Katherine A High is a professor of paediatrics at the University of Pennsylvania, and Director of Research in the Hematology Division of the Children's Hospital of Philadelphia. She is a member of the Scientific Advisory board of Avingen Inc—a company in which she holds equity. “There are more things in heaven and earth, Horatio, than are dreamt of in your philosophy…” William Shakespeare, Hamlet Gene therapy is a novel area of therapeutics in which the active agent is a nucleic-acid sequence rather than a protein or small molecule. The possibilities for gene therapy are enormous. The approach could be used to supply a normal copy of a missing or defective gene in genetic diseases; to improve the safety or efficacy of other drugs; or to restore normal physiology in the setting of acquired disease. These possibilities have been limited, however, by the practical difficulties that have attended efforts to achieve efficient gene transfer and gene expression for an appropriate duration, and much of the first decade of clinical gene therapy has been characterised by disappointments arising from inefficient gene transfer, only transient gene expression, or both. As technology related to gene-delivery vehicles has evolved, clinical results have moved closer to the desired outcome. A gene-therapy approach to treatment of any disease requires three elements: a therapeutic gene to be transferred; a vector to effect the transfer; and appropriately chosen target cells in the recipient. In genetic diseases, the identity of the therapeutic gene is generally not in question, and the first convincing successes have come in these disorders. Clinical gene therapy began with the introduction via a retroviral vector of the gene encoding adenosine deaminase into the lymphocytes of two young patients with severe combined immunodeficiency disease (SCID). Although the treatment had limited success, and the efficiency of gene transfer was quite low, this attempt helped to define the technical limitations of the approach and to establish areas for research. Several conceptual and technical advances have facilitated the successful application of this approach, in which haemopoietic cells are removed from the patient, transduced with the vector bearing the donated gene, and then reinfused: the development of reliable techniques for isolating CD34-positive cells; the characterisation of cytokines that can promote proliferation (required for retroviral transduction) of haemopoietic cells without compromising their reconstitution capacity; the ability to concentrate vector to achieve high-titre preparations; and the definition of optimum conditions for direct contact between the retroviral particle and the target. All these advances contributed to the successful treatment of children with X-linked SCID. This disorder is due to a mutation in the gene encoding the γc cytokine receptor subunit of the receptors for various interleukins that are critical for the growth and differentiation of lymphoid cells. The investigators used a retroviral vector expressing γc, transduced CD34-positive cells from affected children, and found after reinfusion of the transduced cells reconstitution of T, B, and natural killer cells, that continued for over a year. In haemophilia B, adeno-associated viral vectors expressing human factor IX have been introduced into skeletal muscle, and successful gene transfer and expression have been shown on muscle biopsy. Further dose-escalation studies are needed to identify a safe dose at which all patients have a demonstrable therapeutic effect. A plasmid-based approach for treatment of haemophilia A has also shown promise in early-phase clinical testing. In cancer therapy, a modified adenoviral vector that replicates specifically in cells lacking functional p53 has been shown in large trials to induce tumour-specific lysis in head and neck tumours. Moreover, this gene-based approach is synergistic with conventional chemotherapy and radiotherapy. An unresolved question is whether the vector alone would have similar therapeutic effects related to the immune response. One of the lessons of this first decade of clinical gene therapy is the importance of focusing on diseases that can be targeted with the available technology. Thus, in the case of X-linked SCID, the transduced cells have a marked proliferative advantage in vivo compared to untransduced cells and, therefore, engraftment is possible even without prior myeloablation. Continued improvements in gene delivery systems will allow the extension of gene therapy approaches to an ever-increasing range of human diseases. Proof-of-concept studies in animal models have already been completed for retinitis pigmentosa and for lysosomal storage disease, disorders for which there is currently no treatment. Recombinant proteins once appeared far too complex and difficult to standardise, yet they are now a standard part of the therapeutic armamentarium. Similarly, vehicles of gene delivery, whether viral or non-viral, represent a new level of complexity in therapeutics, but recent data suggest that these too can be successfully developed for treatment of human disease.