Glycosylated hemoglobin is an indicator of long-term glycemic control and a strong predictor of diabetic complications. This paper provides a comprehensive description of glycemic control (total glycosylated hemoglobin (GHb)) up to 4.5 years duration of diabetes by age, duration, and sex in a population-based cohort (n = 507) aged less than 20 years followed from diagnosis of Type 1 diabetes in Wisconsin during 1987-1994 Important aspects of demographics and diabetes care are described to allow comparison with other populations. Since large variations between laboratories are known to exist in the measurement of GHb, levels are also interpreted relative to the frequency of short-term complications. GHb increased after diagnosis, but leveled off after 2-3 years. Peak GHb values occurred in the age group 12-15 years. The within-individual standard deviation in GHb between tests, adjusted for age and duration was 1.6%. The mean GHb at last testing was 11.3%, with a standard deviation across individuals of 2.9%. The majority (74%) of individuals saw a diabetes specialist at least once. The mean number of insulin injections per day increased from 2.2 to 2.5 across the 4.5-year duration, and the insulin dose increased from 0.6 to 0.9 units per day per kg body weight. Despite the quite satisfactory level of care, 38% of subjects had GHb levels associated with significant short-term complications.