An objective of The Kroc Collaborative Study Group trial of the feasibility of maintaining improved control of plasma glucose concentrations with continuous subcutaneous insulin infusion (CSII) was to test the non-glycemic aspects of metabolic control in relation to microvascular disease. Serum lipid levels were assessed in the 68 patients completing the 8-mo trial, before and after randomization to conventional insulin treatment (CIT) or CSII. During CSII, fasting serum cholesterol concentrations, normal at baseline (186 +/- 7 mg/dl), were unchanged at 4 and 8 mo (183 +/- 8 and 186 +/- 10 mg/dl). Fasting serum triglyceride concentrations fell on treatment with CSII (baseline 90 +/- 12 mg/dl, 8 mo 60 +/- 7 mg/dl, P less than 0.01), but were unchanged during CIT (baseline 88 +/- 8 mg/dl, 8 mo 83 +/- 10 mg/dl). Thirty-two patients in three centers had 24-h profiles of intermediary metabolites measured at baseline (0), 4, and 8 mo. Mean 24-h venous blood lactate levels fell during CSII (baseline 1.28 +/- 0.12 mmol/L, 4 mo 0.99 +/- 0.4 mmol/L, P less than 0.05; 8 mo 1.05 +/- 0.11 mmol/L), but blood alanine levels were unchanged. Venous blood 3-hydroxybutyrate fell from 0.12 X /divided by 1.18 mmol/L at baseline to 0.06 X /divided by 1.22 mmol/L at 8 mo during CSII (P less than 0.01), mainly due to decreases at 0400 and 0600 h. Decreases in fasting serum triglyceride levels confirm previous investigations of insulin-dependent diabetic subjects treated with CSII; decreases of venous blood lactate and 3-hydroxybutyrate levels toward normal indicate that these metabolic effects of CSII recognized in short-term studies are sustained over an 8-mo period.