Hypothyroid (thyroid stimulating hormone (TSH)> or =20 mIU/l; N=32) participants in the third National Health and Nutrition Examination Survey, Phase 2 (1991-1994) were compared with non-hypothyroid subjects (0.5 mIU/l<TSH<20 mIU/l; N=6490) to examine the relationship between hypothyroidism and hyperhomocysteinemia (serum total homocysteine>12 micromol/l) and hypercholesterolemia (serum total cholesterol>6.2 mmol/l). After controlling for age, gender, and race ethnicity, the odds ratios (95% confidence interval (CI)) relating hypothyroidism to hyperhomocysteinemia and high total cholesterol were 4.9 (1.8-14.0) and 8.0 (2.9-21.9), respectively. Based on 26 hypothyroid and 5811 non-hypothyroid subjects with triglyceride concentration < or =2.82 mmol/l, the odds ratio for the relationship between hypothyroidism and high low-density lipoprotein (LDL)-cholesterol (>4.6 mmol/l by the Friedewald equation) was 5.3 (95% CI, 1.3-20.9). Adding additional terms to the multivariate logistic regression model had little effect on the odds ratios relating hypothyroidism to high total or LDL-cholesterol, but adding terms for serum creatinine concentration >123.8 micromol/l and for red blood cell folate and serum vitamin B-12 concentrations resulted in an attenuated, but still significant (P<0.05), odds ratio relating hypothyroidism to hyperhomocysteinemia (2.5; 95% CI, 1.0-6.1). Controlling for cigarette smoking, heart attack/stroke history, body mass index, and serum albumin concentration did not affect the odds ratios. Hyperhomocysteinemia and hypercholesterolemia could help to explain the increased risk for arteriosclerotic coronary artery disease in hypothyroidism.