Aims/hypothesis. Induction of tolerance to insulin is achievable in animal models of Type I (insulin-dependent) Diabetes mellitus by oral treatment with this hormone, which can lead to prevention of the disease. In the Diabetes Prevention Trial of Type I diabetes (DPT-1), oral insulin is given with the aim of preventing disease insurgence. We investigated whether if given at diagnosis of Type I diabetes in humans, oral insulin can still act as a tolerogen and therefore preserve residual beta-cell function, which is known to be substantial at diagnosis. Methods. A double-blind trial was carried out in patients (mean age ± SD: 14 ± 8 years) with recent-onset Type I diabetes to whom oral insulin (5 mg daily) or placebo was given for 12 months in addition to intensive subcutaneous insulin therapy. A total of 82 patients with clinical Type I diabetes ( < 4 weeks duration) were studied. Basal C peptide and glycated haemoglobin were measured and the insulin requirement monitored every 3 months up to 1 year. Insulin antibodies were also measured in 27 patients treated with oral insulin and in 18 patients receiving placebo at the beginning of the trial and after 3, 6 and 12 months of treatment. Results. The trial was completed by 80 patients. Overall and without distinction between age at diagnosis, at 3, 6, 9 and 12 months baseline mean C-peptide secretion in patients treated with oral insulin did not differ from that of those patients treated with placebo. In patients younger than 15 years a tendency for lower C-peptide values at 9 and 12 months was observed in the oral insulin group. Insulin requirement at 1 year was similar between the two groups as well as the percentage of glycated haemoglobin. Finally, IgG insulin antibodies were similar in the two groups at each time point. Conclusion/interpretation. The results of this study indicate that the addition of 5 mg of oral insulin does not modify the course of the disease in the first year after diagnosis and probably does not statistically affect the humoral immune response against insulin. [Diabetologia (2000) 43: 1000–1004]