Retroviral transduction of hematopoietic stem cells (HSCs) offers an attractive strategy for treating malignancies that home to the marrow. This approach should therefore be of interest for evaluating the therapeutic activity of anti-angiogenic agents on hematopoietic malignancies whose growth has been associated with enhanced angiogenesis. A variety of studies have indicated endostatin to be a potent anti-angiogenic agent both in vitro and in vivo, and a human malignancy that might be sensitive to endostatin is human B-lineage acute lymphoblastic leukemia (B-ALL). The demonstrated ability of human B-ALL cells to engraft the marrow of immunodeficient mice suggested the potential of this system for testing an endostatin delivery strategy using co-transplanted non-obese diabetic-scid/scid (NOD/SCID) HSCs engineered to express endostatin. Here we show that, in spite of their mutant scid gene, NOD/SCID HSCs can be transduced with an endostatin-encoding retrovirus at efficiencies that result in a several-fold increase in endostatin serum levels in transplanted recipients. However, this did not alter the regrowth of co-transplanted human B-ALL blasts. These findings validate this gene transfer approach for investigating effects of novel therapeutics on primary human malignant cells that engraft NOD/SCID mice and question the utility of native endostatin for controlling human B-ALL in vivo.