BACKGROUND: Abdominal aortic aneurysm diameter is usually measured by the maximum minor-axis diameter on axial computed tomography (CT). However, this "traditional" diameter may underestimate the real size, as the aorta is not always straight and the aneurysm shape is sometimes in the form of an ellipse along the cross section. Therefore, we measured maximum major-axis diameters using a three-dimensional (3D) workstation and compared them with the traditional maximum minor-axis diameters measured using thin-slice axial CT. METHODS: CT data of 141 AAA patients (with fusiform aneurysms) were stored in a 3D workstation. These thin-slice CT images were reviewed on the 3D workstation to obtain curved multiplanar reconstruction images (CPR images). Using the CPR images, we measured the maximum major-axis and minor-axis diameters on CPR and the angle of the aneurysms to the body axis. RESULTS: The mean traditional maximum minor-axis diameter was 51.2 ± 8.2 mm, whereas the mean maximum major-axis diameter on CPR was 54.7 ± 10.1 mm. Sixty eight patients had a mean aneurysm size of <50 mm when measured by the traditional minor-axis diameter. Among these patients, five (7.4%) had a major-axis diameter >55 mm on CPR. CONCLUSION: The measurement of the traditional maximum minor-axis diameter of aneurysms is useful in the case of most patients. However, the traditional maximum minor-axis diameter may underestimate the real aneurysmal diameter, particularly in patients with an ellipse-shaped aneurysm. The maximum major-axis diameter as measured using CPR images is effective for representing the real aneurysmal size.