Glycaemic control on pump versus pen treatment was evaluated and the effects of optimised metabolic control on kidney function was studied in very long-term uncomplicated insulin-dependent diabetes mellitus (IDDM). Ten otherwise healthy patients participated, age: 36.5 yr +/- 7.9, diabetes duration: 23.7 yr +/- 2.9, urinary albumin excretion (UAE): 5.8 micrograms/min x/ divided by 2.2, se-creatinine and blood pressure were normal and only background retinopathy was present. A 2 x 6 months randomised cross-over study was performed using continuous subcutaneous insulin infusion (CSII) and multiple injection technique (MIT). Glycaemic control was evaluated by a six point profile every two weeks and by measuring HbA1c monthly. At 0, 6 and 12 months, glomerular filtration rate (GFR) and renal plasma flow (RPF) were measured by the constant infusion technique, and UAE by radioimmunoassay. Glycaemic control was significantly better on CSII as compared to MIT (p = 0.01) or pre-study conventional treatment (CT), p = 0.03, whereas there was no difference between MIT and CT. There was no change in kidney function during either treatment. Thus, in these very long-term uncomplicated patients, glycaemic control was significantly improved during CSII. In spite of this, no change was found in GFR, which might suggest that in long-standing diabetes, kidney function is unaltered by changes in metabolic control.