OBJECTIVE We have recently demonstrated multiple aberrations in the GH–IGF axis in the sera of children with untreated insulin‐dependent diabetes mellitus (IDDM) which were restored after insulin replacement. However, the net result of these alterations in the IGF system on the concentrations of free/biologically available IGF‐I in the serum have not been examined directly in diabetic children. In the present study, the effect of diabetes and subsequent insulin replacement on the circulating free IGF‐I concentrations are assessed.DESIGN Fasting venous serum samples were obtained longitudinally, before and at various times after the initiation of insulin treatment in untreated diabetic subjects.SUBJECTS Ten prepubertal, aged (mean ± SEM) 6.3±1.0 years, and six adolescent, aged 12.7±1.1 years, subjects with newly diagnosed and untreated IDDM, and age and pubertal status‐matched control children and adolescents were recruited.METHODS The serum samples were collected before initiating insulin treatment and 12–24h, 1 week, and 1 month thereafter in subjects with IDDM. Insulin doses ranged from 0.5 to 1.2 U/kg/day.MEASUREMENTS Free IGF‐I concentration was assayed by a recently developed two‐site immunoradiometric assay. Total IGF‐I was measured by radioimmunoassay after acid–ethanol extraction of binding proteins. Differences in free and total IGF‐I concentrations in IDDM subjects before and during insulin treatment were analysed by repeated measures analysis of variance followed by pairwise multiple comparisons test. In seven subjects with IDDM, where serum IGFBP‐1 and IGFBP‐3 concentrations, and IGFBP‐3 protease activity had also been measured in a previous study, the relationship between these variables and circulating free IGF‐I concentrations were examined by linear regression analysis.RESULTS Free IGF‐I concentrations in prepubertal subjects with IDDM were 0.9±0.2, 1.5±0.3, 1.6±0.3 and 2.5±0.4μg/l before, 1 day, 1 week and 1 month after insulin treatment, respectively. Free IGF‐I concentrations of control prepubertal children were 2.6±0.5μg/l. Pubertal subjects had higher free IGF‐I concentrations than prepubertal subjects but demonstrated a similar type of pattern; before insulin 2.3±1.1, 1 day 3.8±1.3, 1 week 3.7±0.6, 1 month 6.5±1.5vspubertal controls 7.7±2.0μg/l. Total IGF‐I concentrations were also reduced in untreated diabetic subjects and showed a slower pattern of normalization than free IGF‐I concentrations. Free IGF‐I concentrations correlated positively with total IGF‐I and negatively with IGFBP‐1 concentrations. There was no significant correlation between free IGF‐I and either serum IGFBP‐3 concentrations or IGFBP‐3 protease activity.CONCLUSION Alterations in the IGF system during untreated IDDM lead to a reduction in circulating free IGF‐I concentrations which is restored progressively during insulin treatment. An increase in free IGF‐I precedes that of total IGF‐I suggesting that the former is a more sensitive indicator of the metabolic status. An inverse correlation between free IGF‐I and IGFBP‐1 supports the h