A small number of blood transfusions (1-3) seems sufficient to improve the cadaveric renal allograft outcome, probably via induction of some nonspecific suppressive activity. This activity was assessed by the concanavalin A (Con A) enhancement method; when present, the response of freshly isolated patients' cells to a submitogenic dose of Con A was lowered, leading to a Con A ratio greater than 5, significantly (P less than 0.0001) higher than the one observed in normal controls or untransfused uremic patients. The correlation between this suppressive activity and graft outcome was determined. Thirty-five patients were studied over a 12-month period for graft function (creatinine level) and survival. Both parameters were significantly improved in the group of patients whose Con A ratio was greater than 5 after transfusions. A soluble suppressor factor, or factors, released into the supernatant of patients' lymphocytes cultured for 48 hr, seems responsible for this suppressive activity. Moreover this process is nonspecific, since it suppresses mitogenic response of cells isolated from normal untransfused volunteers, and could be observed when peripheral blood mononuclear cells were used, but not with purified adherent or nonadherent cells. Addition of indomethacin to the cells during the elaboration of the supernatant abolished this activity. However, amounts of PGE2 secreted into the supernatant during the 48 hr of culture could not be correlated with this suppressive activity. These findings suggest that induction of nonspecific immunosuppression by a few blood transfusions could predict a better kidney graft outcome.