The aims of this survey were (1) to estimate the prevalence of known diabetes mellitus in 1988 in Casale Monferrato (Northern Italy); (2) to validate different data sources available in Italy; (3) to identify a population-based cohort of diabetic patients. Multiple independent data sources were used and the capture-recapture method was applied to estimate the completeness of ascertainment of the survey. The primary data source was the list of all patients attending the diabetic clinic or those referred by family physicians and paediatricians of the area. The secondary data sources were the list of hospital discharges, the prescriptions data source and the list of all people using reagent strips and insulin syringes. On 1 October 1988 (the cut-off date) 2,069 cases of known diabetes were identified. The estimated completeness of ascertainment of the survey was 91%. Prevalence of known diabetes, Type 1 (insulin-dependent), Type 2 (non-insulin-dependent) and insulin-treated diabetes were, respectively, 2.21% (95% CI 2.13-2.29), 0.80/1,000 (0.62-0.98) and 2.10% (2.01-2.19), 2.92/1,000 (2.57-3.27). A higher prevalence of Type 2 diabetes was observed in women (2.30%, 2.18-2.42) than in men (1.88%, 1.76-2.00). Age-specific prevalence of Type 2 diabetes increased with age. Computerized data sources routinely available in the Piedmont Region (hospital discharges and prescriptions data sources) showed a low completeness of ascertainment when considered together (65%, 1,338 of 2,069), indicating the need to involve the diabetic clinic and family physicians in the ascertainment of known diabetes. In conclusion, the prevalence of known diabetes in Italy was lower than in Northern Europe and the United States.