The Diabetes Control and Complications Trial was conducted in heavily-equipped centres on a selected and motivated patient cohort. The aim of the present study was to evaluate at one-year follow-up the results of intensive insulin therapy in patients with insulin-dependent diabetes mellitus attending a department of diabetology. From October 1, 1993, to December 31, 1994, all our hospitalised patients under 55 years of age with HbA1c levels above 8% and receiving 2 daily insulin injections were offered the opportunity to shift to 3 daily injections (short-acting insulin in the morning and at midday, and a mixture of short-acting and intermediate insulin in the evening). Patients were instructed to increase blood glucose self-monitoring and to see their diabetologist more often (once every two months). Five patients refused and 45 accepted this proposal: 22 women and 23 men (mean age 31.8 +/- 10.9 yr), BMI 23.5 +/- 2.9 kg/m2, duration of diabetes 12.8 +/- 10.1 yr, HbA1c 10.0 +/- 2.0%. Five patients were lost to follow-up, 2 asked to have their medical file transferred, 3 returned to 2 daily injections, and 5 consulted only once during the year of follow-up. For the 29 patients seen after one-year follow-up, the decrease in HbA1c levels from 10.0 +/- 1.9% to 9.5 +/- 1.8% was not statistically significant. Sixteen patients complained of increased occurrence of hypoglycaemia (3 comas). In routine clinical practice, the prescription of intensive insulin therapy to non-selected insulin-dependent diabetic patients can be associated with a high number of patients lost to follow-up (17% in our study). An increase in the number of daily insulin injections will improve glycaemic control only if self-monitoring and medical surveillance are also intensified. However, many long-term poorly-controlled insulin-dependent diabetic patients are reluctant to comply with these recommendations.