OBJECTIVE: To shorten the period of carbohydrate intolerance preceding the diagnosis of IDDM in children. RESEARCH DESIGN AND METHODS: The incidence of diabetic ketoacidosis (DKA) was studied in newly diagnosed diabetic children aged 6-14 years, in the area of Parma, Italy, 8 years after an information program on DKA was introduced to teachers, students, parents, and pediatricians. Information was provided by displaying a poster with a few practical messages in 177 primary and secondary public schools. The pediatricians working in the same area were given equipment for the measurement of both glycosuria and blood glucose levels, as well as cards listing guidelines for the early diagnosis of diabetes, to be given to patients. A toll-free number was also provided. Clinical and laboratory features of 24 young diabetic patients diagnosed in the Parma area (group 1) were compared with those of 30 patients coming from two nearby areas in which no campaign for the prevention of DKA had been carried out (group 2). RESULTS: From 1 January 1991 to 31 December 1997, DKA was diagnosed in 3 children from group 1 (12.5%) and in 25 children from group 2 (83.0%) (chi 2 = 26.8; P = 0.0001). The three cases of DKA in group 1 were observed in 1991 (n = 1) and in 1992 (n = 2). No patients from the Parma area who had DKA were admitted to our unit after 1992. The duration of symptoms before diagnosis was 5.0 +/- 6.0 and 28.0 +/- 10.0 days (P < 0.0001), in groups 1 and 2, respectively, Metabolic derangements were less severe in patients of group 1 than in those of group 2. Hospitalization for the treatment of overt diabetes and for the teaching of self-management of the disease lasted 5.4 +/- 1.2 days in group 1 and 13.3 +/- 2.4 days in group 2 (P = 0.002). The total cost of the 8-year campaign was $23,470. CONCLUSIONS: The prevention program for DKA in diabetic children aged 6-14 years, carried out in the Parma area during the last 8 years, was successful. Thanks to this program, cumulative frequency of DKA in new-onset IDDM decreased from 78% during 1987-1991 to 12.5% during 1991-1997. None of the newly diagnosed diabetic children aged 6-14 years and from the Parma area were ever admitted to the hospital for DKA after 1992.