The search for genes related to the cause of common complex disorders such as cardiovascular disease has been frustrating, partly because of the many factors known to contribute to cardiovascular disease and the potential "distance" of cardiovascular disease as a phenotype from genes and gene products. Linkage and association studies for phenotypes more proximal in the pathway from DNA sequence variation to overt clinical disease, such as ultrasound-defined carotid atherosclerosis, may potentially be more enlightening. Only one genetic variant previously reported to be associated with atherosclerosis or clinically evident cardiovascular disease, matrix metalloproteinase (MMP) 3, has shown consistently positive associations with carotid disease, although it has not been studied widely. Another, PON1 L55M, is weakly associated in subgroups only, and 2, ApoE and MTHFR, are equivocal. Genetic variants reported to be associated with clinical cardiovascular disease show weak or no relationship to carotid atherosclerosis. This may reflect the known inconsistency in associations of genetic variants with clinical cardiovascular disease itself. Alternatively, genetic determinants of ultrasound-defined carotid atherosclerosis may differ from those of clinically manifest cardiovascular disease and may require pursuit through large-scale genomic studies of carotid atherosclerosis as a distinct phenotype. Only 1 genetic variant, MMP 3, has shown consistently positive associations with ultrasonographic carotid disease, although it has not been studied widely. Another, PON1 L55 mol/L, is weakly associated in subgroups only. Genetic variants reported to be associated with clinical cardiovascular disease show weak or no relationship to carotid atherosclerosis.