The presence of subclinical cardiovascular disease has been documented to indicate high coronary risk. This study investigated the impact of subclinical cardiovascular disease derived from echocardiography and carotid ultrasonography on traditional coronary risk stratification using the Framingham risk score (FRS) in a community-based, multiethnic population. Echocardiography and carotid ultrasonography were performed in 1,445 subjects (aged >39 years; 40% men; 53% Hispanic, 20% white, 24% black) from the Northern Manhattan Study. Subclinical cardiovascular disease was defined as the presence of left ventricular hypertrophy and/or carotid plaque greater than the gender-specific 75th percentile of the left ventricular mass index and maximal carotid plaque thickness distribution. The prevalence of subclinical cardiovascular disease was examined in each FRS category (low, intermediate, and high risk). In subjects with low or intermediate FRSs, 35% had subclinical cardiovascular disease (low FRS 29%, intermediate FRS 42%). In the intermediate FRS category, subclinical cardiovascular disease was significantly more prevalent in women than in men (53% vs 32%, p <0.0001) and in black and white subjects than in Hispanics (59% and 46% vs 33%, p <0.0001 and p = 0.040, respectively). In conclusion, the ultrasound assessment of subclinical cardiovascular disease may help reclassify 1/3 of subjects with low or intermediate FRSs into higher risk groups. In the intermediate FRS category, FRS appears to underestimate the coronary risk more in women than in men and more in whites and especially in blacks than in Hispanics.