In the context of allogeneic bone marrow transplantation, an accurate estimate of the risk of developing graft-versus-host disease (GVHD) is of major interest. The pre-transplant frequency of donor's helper T-lymphocyte precursors (HTLp) directed against host's antigens may be helpful in predicting this risk. This technique relies on an indirect measurement of interleukin-2 (IL-2) secreted by the HTLp, as assessed by the proliferation of an IL-2 dependent cell line. Many authors use the murine CTLL-2 cell line in this assay, but these cells do not respond to the presence of minute amounts of IL-2 in the culture medium, and thus do not discriminate between the absence or the presence of very low levels of IL-2. We therefore decided to compare CTLL-2 with another IL-2 dependent cell line, the murine A9.12 cell line. A comparison was made using serial dilutions of recombinant human IL-2, limiting dilutions of baby hamster kidney (BHK) cells transfected with human IL-2 gene and in the context of clinical tests performed for the detection of pre-transplant HTLp. Both the sensitivity and reliability of the tests were better using A9.12. We conclude that the A9.12 cell line might be a more suitable tool for pre-transplant HTLp determinations before allogeneic bone marrow transplantation or whenever low IL-2 levels are to be measured.