In epidemiological studies on the health effect of methylmercury (MeHg) exposure, maternal-hair mercury concentration has been used as an exposure biomarker because of its ease of collection and capability to recapture the exposure history. However, artificial hair-waving reduces the mercury concentration and there is little agreement about which part of the hair strand properly represents fetal exposure. We presented an overview of the studies addressing umbilical cord and mercury in PubMed and evaluated the usefulness of umbilical cord mercury concentrations as biomarkers of fetal exposure to MeHg. The mean total mercury (T-Hg) concentration in cord blood was between 0.5 and 35.6microg/l, and the cord blood-to-maternal blood ratio of T-Hg concentrations was estimated to be approximately 1.5. MeHg concentrations in dried cord tissue did not exceed 0.4 microg/g in Japanese populations without particular exposure to MeHg. Dried cord tissue appeared to be better than wet tissue because the definition of wet weight of the umbilical cord is ambiguous. Both cord-blood and cord-tissue mercury concentrations seemed to correlate closely with maternal-hair and maternal-blood ones. Since cord mercury concentrations are a direct exposure biomarker of the fetus and the cord blood-to-maternal blood ratio of mercury differed markedly among mother-child pairs, mercury concentration in cord blood or dried cord tissue should therefore be used in assessing the possible effects of fetal exposure to MeHg on the susceptible brain. Further studies are required to clarify at which period of exposure during gestation the cord mercury concentration represents in relation to mercury concentrations in maternal segmental hair.