Résumé : The aim of this study was to determine the variability of various measurement protocols for measurement of abdominal aortic aneurysm (AAA) and the clinical relevance of variability. Three radiologists performed computed tomographic angiography measurements of both the aorta and the largest portion of the aneurysm on selected axial slices. Then measurements of the largest portion of the aneurysm were performed on unselected axial slices, sagittal and coronal reformatted. Finally, aortic volume was calculated. Measurements and volume calculation were performed before and after endovascular repair and assessed: Part 1: interobserver variability for maximum anteroposterior (MAP) and maximum transverse (MTR) diameters on selected slices; part 2: interobserver variability for unselected slices considering MAP and MTR; part 3: interobserver variability considering maximum diameter in any direction (MAD); part 4: interobserver variability for sagittal (SAG) and coronal (COR) free curved multiplanar reformation (MPR); and part 5: volume calculations. We then determined which technique of measurement was the most clinically relevant for detecting changes in aneurysm size or aortic volume. Parts 1 and 2: interobserver variability was 4.1 mm for both MAP and MTR; part 3: interobserver variability was 7 mm for MAD; part 4: interobserver variability was 5.5 mm (COR) and 4.9 mm (SAG); part 5: interobserver variability for volume was 5.5 ml. A combination of MAP and MTR was the most useful for detecting aortic modification. Volume calculation was needed in only a few cases. We recommend avoiding MAD and MPR measurements and suggest instead measuring both maximum anteroposterior and maximum transverse diameters. If aneurysm size remains stable after endovascular repair, aneurysm volume should be measured.