Summary. The development of diabetic glomerulopathy in kidneys transplanted to diabetic patients was estimated in transplant biopsies and evaluated in relation to suspected clinical risk factors for diabetic nephropathy. Surgical biopsies were taken at baseline and at 24-36 months post-transplantation in 16 Type 1 (insulin-dependent) diabetic patients and 8 non-diabetic control subjects with a glomerular filtration rate more than 30 ml. min * at follow-up. Immunosuppressive therapy included cyclosporine in all but one case. Stereological methods were used to assess basement membrane thickness, volume fraction of mesangium per glomerulus, and volume fraction of matrix per mesangium. The volume fraction of interstitial tissue per cortex was estimated by light microscopy. After 2 years the basement membrane thickness had increased by 55 nm (SD 58 nm) in the diabetic group. This change was significantly different from that of 2 nm (SD 37 nm) in control subjects (p = 0.02). Mesangial volume fraction increased significantly by 0.04 (SD 0.03) in diabetic patients, and this change was significantly different from that of -0.01 (SD 0.04) in non-diabetic patients (p =0.009). No change was detectable in the matrix expressed as fraction of mesangial volume. An increase in interstitial volume fraction from baseline to 2 years was observed, but was significant only in the diabetic group (p = 0.04). The changes in structural parameters did not correlate with mean values during follow-up of glycated haemoglobin or estimated protein intake, nor was any pattern discernible in the relationship to graft tissue types. The observed increase in basement membrane thickness corresponds to that observed in native kidneys during the first years of diabetes, whereas an increase in mesangial volume fraction-using a different protocol-was not observed in the early phase of the natural development. Absence of correlation with the various risk factors may reflect an irrelevance of these variables within the current range, or their influence may be offset by stronger mechanisms in the transplant situation, and therefore does not appear in this relatively small series.