The importance of glycemic control for the avoidance of micro- and macroalbuminuria after a diabetes duration of 20 years or more was studied in type I (insulin-dependent) diabetic patients diagnosed before an age of 31 years. Glycemic control was assessed as glycated hemoglobin for 5 years or more, on average 9.8 years with 3.2 measurements per year. Of the 197 included patients, 97 men and 100 women, 112 were normoalbuminuric, 45 had microalbuminuria (20-200 mg/L), and 40 had macroalbuminuria (persistent albuminuria > 200 mg/L). Normoalbuminuric patients had a mean HbA1c +/- SEM for the whole measurement period of 6.9 +/- 0.08%, the microalbuminuric patients 7.3 +/- 0.14% (Mann-Whitney test versus normoalbuminuric group, p = 0.025), and macroalbuminuric patients 7.6 +/- 0.13 (p < 0.001). The patients were on average 45 years old, and had an age at onset of 14 years, without significant differences between the groups. Thirteen patients had regressed from macroalbuminuria to normo- (n = 6) or microalbuminuria (n = 7). Of the normoalbuminuric patients, 43% were male, compared to 44% with microalbuminuria and 73% with macroalbuminuria (chi 2, p = 0.026). Logistic regression, with absence of albuminuria as dependent variable, showed a beneficial impact from lower HbA1c and normal blood pressure (blood pressure < 140/90 mm Hg without antihypertensive treatment). In conclusion, these results suggests that absence of albuminuria after 20 years or more of diabetes is associated with lower HbA1c.