Leukocytotoxic antibodies of varied thermal reactivity have been identified in sera of patients pre- and post-transplant and cold antibodies have a reportedly favorable association with good renal allograft survival. Since leukocytotoxicity assays are usually performed only at room temperature (22 C), we questioned the presumption that both warm and cold leukocytotox-ins are reliably detected at 22 C. We sought to define the patterns of thermal reactivity of leukocytotoxic antibodies in alloimmune sera and to relate these antibodies to survival of renal allografts. In presensitized patients, we found with discriminating incubation conditions (15, 22, and 37 C) that pure warm and/or pure cold reactions constituted about 31% of positive reactions and these were missed with only 22 C incubation. Expression of thermal amplitudes is highly target cell-dependent and all three incubation temperatures are required to allow full expression of antibody reactivity of a serum. In 45% of the long-term allograft recipients, we detected circulating leukocytotoxins and room temperature tests missed 48% of the positive reactions. Presensitization was associated with poor graft survival but patients with warm or cold antibodies had graft survival rates approximating those in patients with no detectable antibody. Post-transplant, patients with circulating antibodies to greater than 10% of a panel had poor graft survival but again, those patients with pure warm or pure cold-reactive antibodies had graft survival rates similar to those patients with no antibody. In the majority of recipients with established allografts, no antibody could be detected but in the minority with detectable antibody, there was a slight predominance of the pure warm or cold antibody, compared with the thermal mixture antibody.