Framingham risk scores (FRS) were validated in a mostly Caucasian population. Evaluation of subclinical atherosclerosis by carotid ultrasound may improve ascertainment of risk in nonwhite populations. This study aimed to evaluate carotid intima‐media thickness (cIMT) and carotid plaquing among Mexican Americans, and to correlate these markers with coronary risk factors and the FRS. Participants (n = 141) were drawn from the Cameron County Hispanic Cohort. Carotid artery ultrasound was performed and cIMT measured. Carotid plaque was defined as areas of thickening >50% of the thickness of the surrounding walls. Mean age was 53.1 ± 11.7 years (73.8% female). Most were overweight or obese (88.7%) and more than half (53.2%) had the metabolic syndrome. One third (34.8%) had abnormal carotid ultrasound findings (either cIMT ≥75th percentile for gender and age or presence of plaque). Among those with abnormal carotid ultrasound, the majority were classified as being at low 10‐year risk for cardiovascular events. Carotid ultrasound reclassified nearly a third of the cohort as being at high risk. This discordance between 10‐year FRS and carotid ultrasound was noted whether risk was assessed for hard coronary events or global risk. Concordance between FRS and carotid ultrasound findings was best when long‐term (30‐year) risk was assessed and no subject with an abnormal carotid ultrasound was categorized as low risk by the 30‐year FRS algorithm. Integration of carotid ultrasound findings to coronary risk assessments and use of longer term prediction models may provide better risk assessment in this minority population, with earlier initiation of appropriate therapies.