In recent years heart transplantation has become an increasingly common therapeutic measure for cases of heart failure not amenable to other forms of management. Despite major advances, graft rejection continues to be the main cause of transplant failure. To expand our knowledge of the rejection process, we have grown lymphocytes from biopsy specimens obtained from potentially rejecting allografts. The growth of the lymphocytes requires the presence of interleukin-2 and was shown to correlate well with histologic criteria in identifying rejection (p less than 0.003, chi-square analysis). Of 125 biopsy specimens, 61 (49%) became positive for lymphocyte growth within 30 days. Of the biopsy specimens that were positive on culture, 54 (88%) were histologically diagnosed as rejecting, and seven (11%) were histologically nonrejecting. Of the 64 specimens that were negative on culture, 28 (44%) were histologically negative for rejection, and 36 (56%) were positive for some grades of rejection. Of these culture negative, microscopically positive specimens, however, 31 (86%) were associated with mild (grade 3) or equivocal (grade 2) rejection. Discrepancies between culture and histology appeared to be related to degree of rejection, quantity of sample, and sampling variation. Phenotypic identification of cultured cells showed that the suppressor/cytotoxic phenotype was the predominant cell type in all cultures (mean +/- standard error = 83% +/- 3% of cells). Antihuman mature inducer/helper T cell (OKT4+) T-lymphocytes in culture were associated with biopsy specimens obtained immediately before onset or at the beginning of rejection. Preliminary studies testing in vitro sensitivities of lymphocytes to antithymocyte globulin (ATG) and other control immunosuppressants not believed to function by direct cytolysis (methylprednisolone, cyclosporine, azathioprine) were undertaken. These yielded reproducible dose-response curves, and the drug concentration producing 50% cytotoxicity (LD50) was calculated for each drug-culture combination. The mean LD50 for ATG was in the range of in vivo therapeutic concentrations, whereas LD50 values for control drugs were threefold to twelvefold higher than therapeutic concentrations. Resistance to the effects of ATG showed a tendency to correlate with prolongation of rejection; however, a similar tendency was noted for control drugs, suggesting the possibility of a nonspecific mechanism of resistance to the effects of ATG. The technique of interleukin-2 lymphocyte culture from rejecting allografts promises to further our understanding of mechanisms of graft rejection.