It is accepted that antithymocyte globulin (ATG) preparations vary in their bioactivity and side effects. However, this is poorly documented in the literature. We compared the clinical course and cytokine response of heart transplant patients who had received either Merieux or Stanford ATG preparations. The serum cytokine response (interleukin [IL]-6, tumor necrosis factor [TNF]-alpha, IL-4, and IL-10) of 28 consecutive heart transplant recipients was measured for 14 days after surgery using ELISAs. The effect of various ATG preparations on cytokine stimulation of whole blood in vitro was also evaluated. There was a much greater in vivo IL-6 and TNF-alpha response to Merieux than to Stanford ATG (P < 0.0005). There was little IL-4 or IL-10 response with either preparation. No side effects could be attributed to either treatment. No significant difference was seen in the frequency of rejection at 30, 90, or 365 days. More infection episodes occurred in the group treated with Stanford ATG at 30 days (0.5 compared with 0.2 episodes/patient; P = 0.097), 90 days (1.2 compared with 0.5 episodes/patient; P = 0.17), and 365 days (2.8 compared with 1.8; P = 0.59), although none of these differences were statistically significant. When tested in vitro for cytokine stimulation, the in vivo pattern was confirmed, with Merieux ATG producing greater levels of TNF-alpha and IL-6 than Stanford ATG. The differences in cytokine stimulation may be reflected in different immunosuppressive activities. Further research to elucidate the important components of immunosuppressive activity while excluding potentially detrimental effects is important.