Aim: To compare 4D flow magnetic resonance imaging (MRI) and 2D phase contrast (PC) MRI when evaluating bicuspid (BAV) and tricuspid (TAV) aortic valves. Materials and Methods: A total of 83 subjects (35 BAV, 48 TAV) were explored with 4D flow and 2D PC MRI. Systolic peak velocity, peak flow and regurgitation fraction were analysed at two pre-defined aortic levels (aortic root, midtubular). Furthermore, the two methods of 4D flow analysis (Heart and Artery) were compared. Results: Correlation between the 2D PC MRI and 4D flow MRI derived parameters ranged from moderate (R=0.58) to high (R=0.90). 4D flow MRI yielded significantly higher peak velocities in the tubular aorta in both groups. Regarding the aortic root, peak velocities were significantly higher in the TAV group with 4D flow MRI, but in the BAV group 4D flow MRI yielded non-significantly lower values. Findings on peak flow differences between the two modalities followed the same pattern as the differences in peak velocities. 4D flow MRI derived regurgitation fraction values were lower in both locations in both groups. Interobserver agreement for different 4D flow MRI acquired parameters varied from poor (ICC=0.07) to excellent (ICC=1.0) in the aortic root, and it was excellent in the tubular aorta (ICC=0.8-1.0). Conclusion: 4D flow MRI seems to be accurate in comparison to 2D PC MRI in normal aortic valves and in BAV with mild to moderate stenosis. However, the varying interobserver reproducibility and impaired accuracy at higher flow velocities should be taken into account in clinical practice when using the 4D flow method.