Many problems obviously continue to exist in gene transfer using retroviruses as a means of inserting foreign DNA into hematopoietic stem cells, especially with regulated genes such as the human beta globin genes. First, it is unclear whether the available retroviral vectors will infect enough stem cells for gene transfer to be successful over the long term. Second, there may be sequences necessary for normal beta globin gene expression that may also inhibit the normal retroviral life cycle, thus decreasing the efficiency of gene transfer or gene expression. It seems clear that in order to optimize the success of gene transfer, the highest possible titer of viral production is necessary. New approaches are aimed at increasing viral titer. The transfect/infect method appears useful. Growth factors may also be useful by increasing stem cell proliferation. Single growth factors may not be sufficient to optimize stem cell cycling. To date, interleukin-3 seems to be the single most useful growth factor, although interleukin-3 with interleukin-6 or other combinations of growth factors including interleukin-1 and granulocyte-macrophage colony-stimulating factor (GM-CSF) appear to have potential. Future work also is required to optimize the number of marrow stem cells needed for successful transplantation. Long-term bone marrow culture and stromal cell cultures may provide new and improved marrow culture conditions for achieving this goal. Improvements in the efficiency of both gene transfer and gene expression are necessary before we can consider the concept of gene transfer for the treatment of various hematologic genetic diseases in humans.