OBJECTIVE--To compare the assessment of severity of coronary artery stenosis by the conventional pathology methods with a method designed to resemble quantitative angiography. DESIGN--31 human hearts harvested at necropsy were fixed by perfusion of the aortic root with 10% formol saline at 120 mm Hg for 24 hours. The right coronary and left anterior descending coronary arteries were transversely sliced every 2 mm and the absolute lumen dimensions plotted against the distance from the coronary ostium. Stenosis figures were calculated by comparing the lumen diameter with the lumen diameters in adjacent normal arterial segments in a manner identical to that used in angiographic measurement. The coronary artery segments were then processed histologically. Stenosis was then remeasured by comparing the lumen diameter with the diameter of the vessel within the internal elastic lamina identified by elastic van Gieson staining. RESULTS--Compared with the method that was analogous to angiography, the pathology method used on histological slides overestimated the degree of stenosis by 25-30%. The lack of concordance between the methods was not a function of the severity of the stenosis. CONCLUSION--When they read necropsy reports in which the severity of coronary artery stenosis is assessed cardiologists should be aware of the discrepancy between clinical and pathological methods.