Several major meta-analyses have concluded that exposure to environmental tobacco smoke (ETS) increases the risk of coronary heart disease (CHD) by about 25% among never smokers. However, these reviews have excluded a large portion of the epidemiologic evidence on questionable grounds and have been inconsistent in the selection of the results that are included. We conducted an updated meta-analysis and critique of the evidence on ETS exposure and its relationship to death from CHD among never smokers. Our focus is on the U.S. cohort studies, which provide the vast majority of the available evidence. ETS exposure is assessed in terms of spousal smoking, self-reported estimates, and personal monitoring. The epidemiologic results are summarized by means of overall relative risks and dose-response relationships. The methodological issues of publication bias, exposure misclassification, and confounding are discussed. Several large studies indicate that spousal smoking history is a valid measure of relative exposure to ETS, particularly for females. Personal monitoring of nonsmokers indicates that their average ETS exposure from a smoking spouse is equivalent in terms of nicotine exposure to smoking less than 0.1 cigarettes per day. When all relevant studies are included in the meta-analysis and results are appropriately combined, current or ever exposure to ETS, as approximated by spousal smoking, is associated with roughly a 5% increased risk of death from CHD in never smokers. Furthermore, there is no dose-response relationship and no elevated risk associated with the highest level of ETS exposure in males or females. An objective assessment of the available epidemiologic evidence indicates that the association of ETS with CHD death in U.S. never smokers is very weak. Previous assessments appear to have overestimated the strength of the association.