The experimental use of the total artificial heart as an interim device before heart transplantation has demonstrated its potential usefulness. Once again there has been a reaffirmation of the basic principle in transplantation that careful selection of patients is a prerequisite for a high likelihood of success. Our experience suggests that advanced degrees of renal failure and liver dysfunction are not likely to be associated with a consistently useful interim support. Although our numbers of patients are quite small, we continue to be concerned by what appears to be the tendency for the development of severe thrombocytopenia not resulting from the use of heparin. Studies of platelet kinetics and deposition are planned for the future. Although no recipient has had direct evidence of thromboembolic events, systemic anticoagulation with heparin has been aggressively applied, and bleeding complications have been common. All Jarvik-7 hearts on which autopsies have been performed have demonstrated grossly evident platelet fibrin thrombi, especially around the inlet and outlet valve housings. It has not been our purpose to investigate the relative role for univentricular or biventricular assist. It is likely that a number of our patients might have benefited from left ventricular assistance alone. It would appear that additional patients are needed before strong conclusions can be formed regarding the potential benefits and risks of the current Jarvik-7 device or any total artificial heart as an interim method of support before heart transplantation.(ABSTRACT TRUNCATED AT 250 WORDS)