It has been questioned whether aiming at near-normoglycaemia by intensified insulin treatment regimens is feasible and safe for the majority of patients with insulin-dependent diabetes. In this study, intensified insulin injection therapy (including blood glucose self-monitoring and multiple insulin injections) based upon a 5-day inpatient group teaching programme was evaluated in Type 1 (insulin-dependent) diabetes mellitus in the centralised health care system of Bucharest. One hundred patients (group A, initial HbA1 12.5%) were followed for 1 year on their standard therapy (individual teaching, no metabolic self-monitoring), and thereafter for 1 year on intensified therapy. Another 100 patients (group B, HbA1 12.3%) were followed for 2 years on intensified therapy. A third 100 patients (group C, HbA1 11.7%) were assigned to a basic 4-day inpatient group teaching programme with conventional insulin therapy (including self-monitoring of glucosuria and acetonuria) and followed for 1 year. Mean HbA1 remained unchanged after standard treatment (group A: 12.8% at 12 months), but decreased during intensified therapy (group A: 10.1% at 24 months; group B: 9.3% at 12 months, 9.5% at 24 months; p less than 0.0001). In group C, no change was found compared to standard treatment (i.e. group A at 12 months). Incidence rates of ketoacidosis were 0.16 episodes per patient per year during standard treatment, 0.01 during intensified treatment (p less than 0.01) and 0.04 in group C (p less than 0.025). Hospitalisation rates were reduced by 60% during intensified therapy and by 40% in group C. Frequency of severe hypoglycaemia was not significantly different between the three treatment regimens. Thus, under the condition that insulin treatment is based upon a structured and comprehensive training of the patient, intensified insulin injection therapy performed as routine treatment of Type 1 diabetes significantly lowers HbA1 levels without increasing the risk of severe hypoglycaemia.