To assess observer variation between calibrated-catheter digital subtraction angiography (DSA) and software-enhanced multidetector computed tomography angiography (CTA) in measuring vessel length prior to endovascular aortic aneurysm repair (EVAR). Thirty patients (25 men; mean age 65 years, range 61-85) scheduled for EVAR underwent CTA in 4x2-mm collimation using advanced vessel analysis software. CTA measurements were performed twice by 2 blinded readers in random order with at least a 4-week interval between readings. Nine patients were found unsuitable for endovascular repair after the CTA, so DSA was performed in 21 patients for morphometric evaluation of the abdominal aorta and the iliac arteries. The following segments were measured: H1 (aneurysm neck), H2 (lower renal artery to distal aspect of the aneurysm), H3 (lower renal artery to aortic bifurcation), and H4a/H4b (lower renal artery to iliac bifurcations). Length measurements on DSA were made by (1) following the catheter path in the aortic lumen and (2) dividing tortuous vessel anatomy into segments and measuring each segment along an idealized centerline. Addition of the various segments allowed comparison with data obtained from CTA measurements. CTA was performed with good intraobserver agreement for all length parameters except H3 in reader 2 (p<0.05). While good interobserver agreement was demonstrated for CTA over long aortoiliac distances (H4a, H4b), higher interobserver agreement was obtained with DSA for shorter segments (H1, H2). Considerable differences were observed between CTA and DSA for the lengths H2 and H4b. CTA produces better intra and interobserver correlations in measuring vessel length than DSA. It has the potential to replace DSA as an imaging method before EVAR.