There is no doubt that currently available biventricular pneumatic pulsatile devices placed orthotopically with transcutaneous drive lines can support life in patients who may then be successfully transplanted. For the most part, the world experience is with the Jarvik hearts. The current model of choice is the 70-cc device because it is smaller and may be implanted with less fear of "fit problems" than the 100-cc model. As Case 4 illustrates, a fit problem may cause a fatal outcome. But if special precautions in implantation are taken, these may be avoided (Case 6). The fears of excessive bleeding, hemolysis, embolism, and infection with the use of these devices are not as prohibitive as we suspected. Bleeding is always a threat in complex cardiac surgery with grafting. Hemolysis is not a problem if excessive transfusion can be avoided. Embolism and infection may be inevitable, but as our patient B.C. has proven, there is no need to rush immediately to transplantation. And if the implanted patient does not meet local transplant selection criteria, we have enough information now to recommend that they not be accepted for transplantation. Bleeding may be anticipated to be a major problem in any patient with dense reactive scar tissue. To avoid this, transplantation should be done within 3 weeks. We recommend patients who have previously been selected for orthotopic transplantation and begin an accelerated decompensation (our Cases 2 and 7) as the best candidates for temporary orthotopic mechanical support. In these patients, there is no question about whether recovery of the native heart is possible or whether a reversible myocardial insult is present. The plan to do an orthotopic transplant indicates that a cardiectomy is necessary. Case 5, who was in reasonable shape for transplantation, was also favorable. In cases of anticipated graft failure, early use of the Jarvik 7-70 is recommended. This type of device, when suitably placed, provides excellent control of the circulation. There is no requirement for intensive care of the native heart, since it is gone, along with toxic antiarrhythmic medications, risk of embolizing a mural thrombus, and the constant balancing of a univentricular device vis-á-vis the native heart. Further, if pulmonary edema is present with accompanying elevation in pulmonary vascular resistance, the biventricular device requires only an upward adjustment in right drive pressure. With a univentricular device one must worry not only about the pulmonary vascular resistance but also about the capacity of the right heart to pump at a normal output.(ABSTRACT TRUNCATED AT 400 WORDS)