The effects of exposure to environmental chemicals, such as heavy metals and fine particulates, on cardiovascular diseases have been reported. To set a permissible exposure standard, an exposure-response relationship should be elucidated. Thus, epidemiologic evidence in relation to long-term, low-level exposure to environmental chemicals (i.e., lead, cadmium, mercury, arsenic, PM_2.5, carbon disulfide) was reviewed. As a result, there exists a clear exposure-response relationship between exposure to lead or PM_2.5 and cardiovascular diseases, but those cohort studies were carried out in the US only. From epidemiologic viewpoint, “consistency” of the observed relationship must be clarified, especially findings of a study in the Japanese population. It is well known that the distribution pattern of cardiovascular risk factors are quite different between two countries, which could modify the true relationship. This will require studies of large samples from the general Japanese population with control for potential confouders. Moreover, to detect a small effect of exposure within low-level range, random misclassification of exposure as well as that of effect must be minimized. Thus, for studies of heavy metals, sensitive biomarkers identified in toxicological studies would be applied. For PM_2.5, a wide range of population samples should be covered by a national sampling network in the near future. Finally, the integration of epidemiologic and toxicological evidence would be more important for risk assessment of environmental chemicals at low exposure levels, because an exposure-response relationship from only epidemiologic observation may be unstable with a wide confidence interval around the threshold.