Some retrospective nonrandomised or cross-sectional studies have shown that higher blood glucose levels are associated with more pronounced microvascular complications in patients with insulin-dependent diabetes mellitus (IDDM). The prospective randomised studies had, until recently, been less definitive. Intensified treatment, and thus lower blood glucose levels, has led to an initial worsening of retinopathy, but this tendency towards more advanced retinopathy has been transient. Albuminuria and manifest neuropathy have been retarded to some extent. Today, 2 long term randomised studies, the Stockholm study and the Diabetes Control and Complications Trial (DCCT), have proven that a lowering of mean blood glucose levels, measured as a lower glycosylated haemoglobin (HbA1c) value, retards or halts retinopathy, nephropathy and peripheral neuropathy. Intensified treatment, whether performed with multiple injections or insulin pumps, leads to some weight gain and a 3-fold increase in the frequency of severe hypoglycaemic episodes. Hypoglycaemia did not cause long term reduced cognitive function in either study, but was unpleasant to the patients. A great majority of patients in the Stockholm study stated that their well-being had increased while participating in the study. The Stockholm programme required 35 minutes extra per patient per month, and a physician and a nurse could tutor 400 patients. This would bring a significant reduction of serious complications and a gain in terms of patient discomfort and cost. A programme of intensified treatment for IDDM is generally indicated and is possible to carry out.