I report on my 6-year experience identifying the correlation of low antithrombin III (AT III) levels with coronary artery disease in a rural Missouri family practice setting. I measured plasma AT III levels in 78 patients who were at risk for coronary artery disease. I also tracked lipid profiles, HBA1C, blood pressure, body mass index, hostility, and nicotine use. I measured AT III levels using Baxter Diagnostic AG Dade antithrombin III chromogenic coagulation assays of antithrombin III (normal range, 88% to 126%) at the Boone County Hospital Laboratory, Columbia, Missouri. I analyzed the patient records and looked for the worst recording of each of the above risk factors, selecting the highest blood pressure, total cholesterol, LDL cholesterol, triglycerides, HBA1C, weight, and age. I selected the lowest AT III level, which often occurred when the patient was stressed. When the worst risk factors were picked over a long period of time, low AT III correlated with an arbitrary coronary artery disease scale 24% of the time. This negative correlation (r = -.49) was higher than any other risk factor in these selected patients.