Long-term consequences of glomerular enlargement after transplantation are not well understood. The aim is to evaluate the relationship between glomerular volume (Vg) estimated in protocol biopsies, graft function and graft survival. Vg and Banff chronic damage score were evaluated in protocol biopsies at 4 months. Creatinine clearance (CrCl) was estimated by the Cockroft-Gault formula. Vg estimated in 144 patients was 4.8 +/- 2.0 x 10(6)mu(3). It was associated with donor age (r = 0.23, p < 0.01), recipient body mass index (r = 0.17, p = 0.04), delayed graft function (Vg = 5.9 +/- 2.3 vs. 4.6 +/- 1.9 x 10(6)mu(3), p < 0.01) and CrCl (r = 0.17, p = 0.04). The best cutoff of Vg, Banff chronic damage score and CrCl was determined by Cox regression analysis, being 5.0 x 10(6)mu(3) for Vg (relative risk (RR): 2.4, 95% confidence interval (CI): 1.03-5.6), >2 for chronic damage score (RR: 3.4, 95% CI: 1.03-8.9) and 60 mL/min for CrCl (RR: 3.5, 95% CI: 1.04-11.9). These variables were independent predictors of death-censored graft survival. According to Vg and CrCl, four groups of patients were defined. Patients with small glomeruli and high CrCl had a 95% graft survival while patients with large glomeruli and low CrCl had a 45% graft survival at 15 years (p < 0.01). Large glomerular volume, high Banff chronic score and poor early renal function in stable grafts are independently associated with death-censored graft survival.