This report deals with 120 cadaveric renal transplants performed in 101 pediatric recipients in this Centre in two five-year periods, 1984 to 1988 (N = 65) and 1989 to 1993 (N = 55). In the first group transplants were allocated on the basis of best size (small donors for small recipients); in the second group priority was given to beneficial HLA matching. Initial immunosuppression was either cyclosporine (CsA) monotherapy (15 mg/kg/day), or triple therapy (CsA 5 mg/kg/day, prednisolone 1 mg/kg/day and azathioprine 1 mg/kg/day) if there was delayed graft function. Patient survival at one year and five years (97.5% and 92.3%, respectively) did not differ between the two groups, although there was an improvement in graft survival at one and five years in the second period relative to the first: 69.2% and 53.8% versus 78.6% and 65.6%. This did not achieve statistical significance. One year graft survival in recipients under five years did not differ significantly from older children (72%). There was a trend to improvement in one year graft survival in the < five years of age pediatric patients in Group 2, with beneficially matched kidneys and improved immunosuppressive management. Graft losses due to acute rejection were similar in both groups. Donor age < 4 years significantly reduced one year graft survival (63% vs. 85%, P = 0.01), while recipient age had no effect. Small donor kidneys were associated with a higher incidence of graft thrombosis. Transplantation resulted in the normalization or acceleration of growth velocity in (84%) of the pre-pubertal children who completed follow up. In conclusion, we have shown that excellent patient and graft survival can be achieved in children transplanted under the age of five years. Kidneys from donors under the age of four years are associated with an unacceptable rate of graft loss. Small children do not readily accept cyclosporine monotherapy. Successful early renal transplantation offers the best chance of normal growth and development.