A total of 196 kidney transplant donors and recipients have been tested with a large panel of antisera and classified as to whether they are matched or mismatched for major leukocyte groups. The survival of the matched patients in comparison to the mismatched patients was then analyzed. During the 21/2 year interval of study, 50 patients have died or have had their rejected kidneys removed. Of these rejected grafts 15 occurred in matched patients and 35 in mismatched patients (P < 0.025). The difference was greater when only retrospectively studied patients who had survived more than 3 months at the time of study were considered. Among 97 patients, 11 deaths have occurred; 10 occurred in mismatched patients whereas only 1 occurred in a matched patient (P < 0.025). The rate of survival in matched recipients was higher than in mismatched patients as shown by the plotting of survival curves (by actuarial methods). Survival after the time of testing was plotted and shown to decline continuously in mismatched patients in comparison to matched patients. An improved technique for the performance of the microdroplet lymphocyte cytotoxicity test is also presented. It is concluded that although survival in this series is associated with matching in marginal limits, many hazards which influence survival especially in the early 3 month period may be making it difficult to demonstrate the association. Further experience, however, will be necessary to rule against transplantation on the basis of the test.