Background: Some authors consider that secondary prevention should be conducted for all DM2 patients, while others suggest that the drug preventive treatment should start or be increased depending on each patient's individual CVR, estimated using cardiovascular or coronary risk functions to identify the patients with a higher CVR. The principal objective of this study was to assess three different cardiovascular risk prediction models in type 2 diabetes patients. Methods: Multicentre, cross-sectional descriptive study of 3,041 patients with type 2 diabetes and no history of cardiovascular disease. The demographic, clinical, analytical, and cardiovascular risk factor variables associated with type 2 diabetes were analysed. The risk function and probability that a cardiovascular disease could occur were estimated using three risk engines: REGICOR, UKPDS and ADVANCE. A patient was considered to have a high cardiovascular risk when REGICOR ≥ 10 % or UKPDS ≥ 15 % in 10 years or when ADVANCE ≥ 8 % in 4 years. Results: The ADVANCE and UKPDS risk engines identified a higher number of diabetic patients with a high cardiovascular risk (24.2 % and 22.7 %, respectively) compared to the REGICOR risk engine (10.2 %). The correlation using the REGICOR risk engine was low compared to UKPDS and ADVANCE (r = 0.288 and r = 0.153, respectively; p < 0.0001). The agreement values in the allocation of a particular patient to the high risk group was low between the REGICOR engine and the UKPDS and ADVANCE engines (k = 0.205 and k = 0.123, respectively; p < 0.0001) and acceptable between the ADVANCE and UKPDS risk engines (k = 0.608). Conclusions: There are discrepancies between the general population and the type 2 diabetic patient-specific risk engines. The results of this study indicate the need for a prospective study which validates specific equations for diabetic patients in the Spanish population, as well as research on new models for cardiovascular risk prediction in these patients. Keywords: Type 2 diabetes, Cardiovascular risk prediction, Cardiovascular disease, Risk prediction models, Primary prediction