Background: Aortic valve area (AVA) estimation in patients with aortic stenosis may be obtained using several methods. This study was undertaken to verify the cardiovascular magnetic resonance (CMR) planimetry of aortic stenosis by comparing the findings with invasive catheterization, transthoracic (TTE) as well as tranesophageal echocardiography (TEE) and anatomic CMR examination of autopsy specimens. Methods: Our study was performed in eight patients with aortic valve stenosis. Aortic stenosis was determined by TTE and TEE as well as catheterization and CMR. Especially, after aortic valve replacement, the explanted aortic valves were examined again with CMR ex vivo model. Results: The mean AVA determined in vivo by CMR was 0.75 +/- 0.09 cm(2) and ex vivo by CMR was 0.65 +/- 0.09 cm(2) and was closely correlated (r = 0.91, p < 0.001). The mean absolute difference between AVA derived by CMR ex vivo and in vivo was -0.10 +/- 0.04 cm(2). The mean AVA using TTE was 0.69 +/- 0.07 with a significant correlation between CMR ex vivo (r = 0.85, p < 0.007) and CMR in vivo (r = 0.86, p < 0.008). CMR ex vivo and in vivo had no significant correlation with AVA using Gorlin formula by invasive catheterization or using planimetry by TEE. Conclusion: In this small study using an ex vivo aortic valve stenosis model, the aortic valve area can be reliably planimetered by CMR in vivo and ex vivo with a well correlation between geometric AVA by CMR and the effective AVA calculated by TTE.