A controlled clinical trial was started to evaluate whether small doses of blood given pretransplant determine a transfusion effect while reducing the risk of antibody production. For this purpose, 65 consecutive never transfused patients suffering from end-stage renal failure were assigned to one of two groups: the first group was transfused with 1 unit of packed red cells (containing a mean of 2350 x 10(6) leukocytes, 900 x 10(6) mononuclear cells) 3 times at 15-day intervals. The second group received one transfusion of about 30 ml of blood adjusted to contain 100 x 10(6) mononuclear cells. While no definitive conclusions are still possible, preliminary data indicate the following: (1) three small transfusions are capable of immunizing the recipient, but lymphocytotoxic antibodies tend to disappear rapidly; (2) in vitro lymphocyte response to lectins of patients receiving small transfusions is not significantly different from that of patients receiving standard transfusions; (3) the two groups of patients differ significantly as far as the T4+ /T8+ cell ratio is concerned: in fact while a decrease of the ratio is observed after standard transfusions, small transfusions determine an increase of the ratio, mainly due to a decrease in the number of T8+ cells; and (4) the clinical course and survival of the graft is worse in patients treated with small transfusions than in those treated with standard transfusions.