Background: Hyperglycaemia impairs tubulo-glomerular feedback. We tested whether variable tubulo-glomerular feedback during hyperglycaemia contributes to renal risk heterogeneity seen in Type 1 diabetes. Methods: During the period 1990-92, we studied the tubulo-glomerular feedback in Type 1 diabetic patients at high or low renal risk [21 of 54 with glomerular hyperfiltration and/or microalbuminuria against 11 of 55 with normal glomerular filtration rate (GFR) and urinary albumin despite uncontrolled diabetes]. The GFR, effective renal plasma flow, mean arterial pressure and fractional reabsorptions of glucose, osmols, sodium and lithium were measured sequentially during normo- and hyperglycaemia. All patients were followed up until 2016 for incident proteinuria, estimated GFR <60 mL/min/1.73 m2, doubling of serum creatinine, end-stage renal disease or all-cause death. Results: Glycaemia increased from 6.1 +/- 1.3 to 15.1 +/- 1.9 mmol/L in both high-risk and low-risk patients. Glycosuria was lower in the high- versus low-risk patients: 0.34 +/- 0.25 versus 0.64 +/- 0.44 mmol/min (P = 0.03). Both groups displayed similar kidney function during normoglycaemia. Hyperglycaemia increased more importantly GFR and fractional reabsorptions, and pre-glomerular vasodilatation in the high- than in the low-risk patients (all P < 0.05). Over 21 years, 31.5% high- versus 12.7% low-risk patients developed endpoints (adjusted P = 0.006). In a multi-adjusted survival analysis of patients having undergone renal tests, each 0.10 mmol/min glycosuria during hyperglycaemia reduced the outcome risk by 0.72 (95% confidence interval 0.49-0.97, P = 0.03). Conclusions: Reduced tubulo-glomerular feedback and glycosuria during hyperglycaemia indicate high renal risk for Type 1 diabetic patients. Inter-individual variability in tubulo-glomerular feedback activity determines renal risk in Type 1 diabetes.< Réduire