While small clinical trials have shown that improved glycemic control reduces the risk of progression of microalbuminuria to proteinuria, two recent clinical trials did not confirm this finding. We sought to reconcile the contradictory evidence by examining the dose-response relationship between hyperglycemia and progression of microalbuminuria to proteinuria in individuals with type 1 diabetes and microalbuminuria (n = 312) who were followed for 4 years with repeated assessments of urinary albumin excretion. Since 33 patients did not participate in follow-up (10.6%), data for 279 patients were analyzed. Urinary albumin excretion level worsened to proteinuria in 40 (4.1 per 100 person-years). To examine the dose-response relationship, baseline HbA1c was divided into four roughly equal groups using the cut points 8, 9, and 10%. The incidence rate varied significantly among the four groups (P = 0.008). Among those with HbA1c <8.0%, the incidence rate of progression was only 1.3 per 100 person-years, while it was 5.1, 4.2, and 6.7 per 100 person-years in the three other groups. We used generalized additive models to examine the dose-response curve using HbA1c as a continuous variable and found that the risk of progression rises steeply between an HbA1c of 7.5-8.5% and then remains approximately constant across higher levels. In conclusion, the results of this study suggest that, in patients with microalbuminuria, the risk of progression to overt proteinuria can be reduced by improved glycemic control only if the HbA1c is maintained below 8.5%. Moreover, below that value, the risk declines as the level of HbA1c decreases.