Heart transplantation may well be the most successful transplantation procedure performed today. One-year patient survival rates now exceed 80%. Many factors are thought to account for differences in outcomes among transplantation centers. No attempt has been made to assess consensus among transplantation program directors regarding the major determinants of patient outcome. In the National Cooperative Transplantation Study we evaluated consensus through a detailed survey of all heart transplantation programs active in the United States in 1988. Of the eligible programs, 104 (91%) returned completed surveys. Data on the medical and surgical determinants of outcomes have been analyzed descriptively. Considerable consensus occurred among program directors about the importance of several factors. For example, over 90% of the respondents felt that heart biopsy should be used as the standard rejection monitoring technique and that left and right heart catheterization should be performed annually with coronary arteriography. Over 60% believed that the availability of a left ventricular assist device for temporary use would also enhance patient outcome. Several approaches were considered to have little beneficial effect on outcome. These included cytoimmunologic monitoring and electrocardiography as standard rejection monitoring techniques. Nearly one half of the respondents opposed steroid discontinuation after transplantation. On several other approaches there was a lack of consensus including the use of heterotopic heart transplantation and conversion from cyclosporine because of renal dysfunction. Consensus conferences are now regarded as a means by which technologic innovations can be evaluated and medical practice guidelines can be set. This analysis suggests that consensus is a useful approach toward assessing medical and surgical practices in heart transplantation.