The recent development of leukemia in three patients following retroviral vector gene transfer in hematopoietic stem cells, resulting in the death of one patient, has raised safety concerns for the use of integrating gene transfer vectors for human gene therapy. This review discusses these serious adverse events from the perspective of whether restrictions on vector design and vector-modified target cells are warranted at this time. A case is made against presently establishing specific restrictions for vector design and transduced cells; rather, their safety should be ascertained by empiric evaluation in appropriate preclinical models on a case-by-case basis. Such preclinical data, coupled with proper informed patient consent and a risk-benefit ratio analysis, provide the best available prospective evaluation of gene transfer vectors prior to their translation into the clinic.