Nonobese diabetic (NOD) mice develop an autoimmune disease with a long prodromal period and constitute a model for investigating the prevention of human type 1 diabetes. Since prophylactic insulin injections reduced the incidence of diabetes in NOD mice, we tested a new prophylactic strategy to prevent diabetes in NOD mice consisting of oral administration of insulin, protected in polyalkylcyanoacrylate nanocapsules from degradation in the gastrointestinal tract. In humans, this form of prophylactic insulin administration would be less constraining than insulin injections. Ninety female NOD mice were randomized at weaning and fed once a week (from 60 to 300 days of age) with insulin nanocapsules (100 U/kg) or empty nanocapsules. Within the group fed with insulin nanocapsules, the incidence of diabetes was reduced (38% vs 75%; P < 0.02), the onset of disease was delayed (P < 0.02), and the severity of lymphocytic inflammation of endogenous islets was reduced (P < 0.03). Although the oral treatment was stopped at 300 days of age, the incidence of diabetes at 360 days remained lower in mice previously fed insulin nanocapsules (P < 0.02). Previous feedings with insulin nanocapsules did not protect against cyclophosphamide-induced diabetes, since final incidence of diabetes (sum of the incidence during the initial 360 days and the further CY-induced incidence) reached the final incidence obtained in mice previously fed empty nanocapsules and treated with cyclophosphamide. Intestinal absorption of insulin nanocapsules was evidenced by HPLC separation of human insulin in NOD sera. During cotransfer, T splenocytes from mice fed insulin nanocapsules were able to reduce the capacity of T cells from diabetic donors to adoptively transfer the disease (P < 0.01). Antigens for islet-cell autoantibodies (ICA) in pancreata from both NOD groups were compared by immunofluorescence with the same ICA-positive human sera to ensure that differences were due to quantitative changes in antigen. These antigens, which could serve as an index of a possibly more extended antigen beta-cell rest, were decreased (P < 0.02) and pancreatic insulin content was reduced (P < 0.05) in mice fed with insulin nanocapsules, suggesting a mechanism of 'beta cell rest'. To summarize, early feeding with insulin nanocapsules reduces diabetes and insulitis in the NOD mouse model that mimics human type 1 diabetes. This may be due both to generation of cellular mechanisms that actively suppress disease and a decrease in antigens which makes beta cells less vulnerable to autoimmune aggression.