PURPOSE: To evaluate the degree of image distortion caused by orthodontic devices during a intracranial magnetic resonance angiography (MRA), and to determine the effectiveness of the 3 dimensional phase-contrast (3D PC). MATERIALS AND METHODS: Subjects were divided into group A (n = 20) wearing a home-made orthodontic device, and group B (n = 10) with an actual orthodontic device. A 3.0T MR scanner was used, applying 3D time-of-flight (TOF) and 3D PC. Two board-certified radiologists evaluated images independently based on a four point scale classifying segments of the circle of Willis. Magnetic susceptibility variations and contrast-to-noise ratio (CNR) on maximum intensity projection images were measured. RESULTS: In group A, scores of the 3D TOF and 3D PC were 2.84 +/- 0.1 vs. 2.88 +/- 0.1 (before) and 1.8 +/- 0.4 vs 2.83 +/- 0.1 (after wearing device), respectively. In group B, the scores of 3D TOF and 3D PC were 1.86 +/- 0.43 and 2.81 +/- 0.15 (p = 0.005), respectively. Magnetic susceptibility variations showed meaningful results after wearing the device (p = 0.0001). CNRs of the 3D PC before and after wearing device were 142.9 +/- 6.6 vs. 140.8 +/- 7.2 (p = 0.7507), respectively. In the 3D TOF, CNRs were 324.8 +/- 25.4 vs. 466.3 +/- 41.7 (p = 0.0001). CONCLUSION: The 3D PC may be a solution method for distorted images by magnetic susceptibility in the intracranial MRA compared with 3D TOF.