Patients with diabetes mellitus (DM), both diagnosed (history of) and undiagnosed (by fasting glucose [FG] only), as well as impaired FG have an increased risk of coronary heart disease (CHD), compared with those with normal FG. Elevations in FG levels, even in normoglycemic subjects (<110 mg/dl), may be significantly related to CHD morbidity and mortality. Improving lipid profiles and blood pressure can decrease both CHD morbidity and mortality in these patients. We evaluated the relation of glucose status to lipid levels, other risk factors, and prevalence of CHD using the 1997 American Diabetes Association diagnostic criteria in a representative sample of United States adults studied in the Third National Health and Nutrition Examination Survey from 1988 to 1994. Impaired FG, diagnosed DM, and undiagnosed DM were more prevalent in older age groups; those > or =65 years had increased prevalence compared with those <50 years old (rate ratios for IFG, DM-FG, and history of DM were 3.5, 4.8, and 10.8, respectively). Glycosylated hemoglobin levels were increased by glucose status. The frequency of known CHD risk factors also increased with worsening glucose status. Age-adjusted CHD prevalence was increased with impaired FG (rate ratio 1.47), DM-FG (rate ratio 1.56), and history of DM (rate ratio 1.72), compared with normal FG. Adjusting for age and other CHD risk factors, hyperglycemia was no longer significantly associated with CHD prevalence. Lipid values, especially high-density lipoprotein cholesterol, hypertension, and other CHD risk factors were more strongly associated with CHD than glucose status. Thus, patients with impaired FG, DM-FG, and history of DM should be considered at higher risk for CHD morbidity and mortality. However, hyperglycemia, per se, does not explain the excess risk. In addition to glucose, lipid profiles and blood pressure should be periodically monitored and appropriate treatment provided to reduce morbidity and mortality from CHD.