OBJECTIVES Under estimating CV risk levels of patients may lead to inappropriate care and unintended clinical and economic consequences. It is not known to what extent Canadian physicians correctly identify the CV risk level of patients who have had previous CV events and/or a diagnosis of diabetes mellitus. METHODS A total of 431 General Practioners from across Canada collected pre-specified data on a case record form (CRF) for CV patients during routine office visits between June 2007 and December 2007. Physicians were asked to assess the risk level of each patient during the completion of the CRF. Patients for whom the physician reported a diagnosis of diabetes mellitus (DM) and/or who had a previous history of myocardial infarction (MI), stroke (S), or peripheral artery disease (PAD) were extracted. Physician assessed risk levels were then compared to this subset of commonly accepted high-risk patients. RESULTS Data on 14,982 patients was obtained. A total of 2822 had a previous history of MI, S, or PAD. Physicians correctly assessed 2540 (90.0%) as high risk, however differences were observed between provinces: BC 81.5%, AB 86.9%, QC 92.9%, ON 90.3%, NS 90.9%, NB 85.3%. A further 1850 had DM and a previous history of MI, S, or PAD. Physicians correctly assessed 1778 (96.1%) as high risk, however regional differences were observed: BC 87.5%, AB 89.6%, QC 98.5%, ON 97.0%, NS 94.4%, NB 97.7%. CONCLUSIONS Despite agreement that patients with a previous history of MI, S, or PAD should be viewed as high risk, and that those with DM and a previous history of MI, S, or PAD are high risk, some physicians still under estimate the CV risk. The reasons for this under assessment and the unintended clinical and economic consequence need to be explored further.