OBJECTIVE An association between reactive oxygen species and diabetic micro- and macrovascular complications has been proposed. In the present study, we have examined the effect of an improved blood glucose control on plasma levels of hydroperoxides in patients with IDDM. RESEARCH DESIGN AND METHODS Subjects included 30 young IDDM patients with microalbuminuria who were randomized to receive either continuous subcutaneous insulin infusion (CSII) by a portable insulin pump (n = 15) or conventional insulin treatment (CIT) (n = 15) for 24 months. Plasma levels of hydroperoxides were measured by the ferrous oxidation with Xylenol Orange, version 2 (FOX2) assay. This method measures total lipid hydroperoxides and, unlike other methods, does not suffer from extraction losses. RESULTS The mean HbA1c level was lower in the CSII group at the end of the study than in the CIT group: (mean [95% CI]) 8.6 (8.1–9.1) vs. 9.6 (9.0–10.3)%, respectively (P < 0.002). The level of plasma hydroperoxides was very similar at the start of the study but was significantly lower in the CSII group compared with the CIT group at the end of the study: 2.9 (2.1–3.7) vs. 4.3 (3.2–5.4) μmol/l, respectively (P < 0.02). In the CSII group, hydroperoxides were reduced by 31% from baseline (P < 0.001), whereas there was no change in levels of hydroperoxides in the CIT group. Mean hydroperoxide levels correlated with mean HbA1c during the study (r = 0.39, P < 0.04). Hydroperoxide levels were associated with the levels of microalbuminuria (r = 0.45, P < 0.02). CONCLUSIONS This study provides support for the hypothesis that hyperglycemia is an important factor in the generation of hydroperoxides, and, thus, reactive oxygen species, in the circulation of IDDM patients.