Quantification of mitral regurgitation (MR) using cardiovascular magnetic resonance can be achieved by three indirect methods. The aims of the study were to determine their agreement, observer variability and effect on grading MR severity. The study comprised 16 healthy volunteers and 36 MR patients. Quantification was performed using the 'standard' [left ventricular stroke volume (LVSV)-aortic forward flow (AoFF)], 'volumetric' [LVSV-right ventricular stroke volume (RVSV)] and 'flow' method [mitral inflow (MiIF)-AoFF]. In healthy volunteers without MR, LVSV was larger than AoFF (mean difference ±SD: 12 ± 6 ml, P < 0.0001). Only small differences were found between LVSV-RVSV (3 ± 6 ml) and MiIF-AoFF (1 ± 5 ml). In patients, mitral regurgitant volumes (MRVs)/fractions (MRFs) were larger (P < 0.0001) using the 'standard' method (90 ± 31 ml/51 ± 11%) compared with the 'volumetric' (76 ± 30 ml/42 ± 11%) and 'flow' method (70 ± 32 ml/44 ± 15%). Inter-observer variability was lowest for the 'flow' and highest for the 'volumetric' method, while intra-observer variability was similar for all three methods. In 29 operated patients with severe MR, MRVs were above the guideline threshold (≥60 ml) in 100, 86 and 83% of the cases, and MRFs were above the threshold (≥50%) in 76, 32 and 48% of the cases, when using the 'standard', 'volumetric' and 'flow' method respectively. In conclusion, the choice of method can affect the grading of MR severity and thereby eventually the clinical decision-making and timing of surgery.