Grading of chronic aortic regurgitation (AR) and mitral regurgitation (MR) by cardiovascular magnetic resonance (CMR) is currently based on thresholds, which are neither modality nor quantification method specific. Accordingly, this study sought to identify CMR-specific and quantification method-specific thresholds for regurgitant volumes (RVols), RVol indexes, and regurgitant fractions (RFs), which denote severe chronic AR or MR with an indication for surgery. The study comprised patients with moderate and severe chronic AR (n= 38) and MR (n= 40). Echocardiography and CMR was performed at baseline and in all operated AR/MR patients (n= 23/25) 10 ± 1months after surgery. CMR quantification of AR: direct (aortic flow) and indirect method (left ventricular stroke volume [LVSV]- pulmonary stroke volume [PuSV]); MR: 2 indirect methods (LVSV- aortic forward flow [AoFF]; mitral inflow [MiIF]- AoFF). All operated patients had severe regurgitation and benefited from surgery, indicated by a significant postsurgical reduction in end-diastolic volume index and improvement or relief of symptoms. The discriminatory ability between moderate and severe AR was strong for RVol >40ml, RVol index >20ml/m2, and RF >30% (direct method) and RVol >62ml, RVol index >31ml/m2, and RF >36% (LVSV-PuSV) with a negative likelihood ratio ≤ 0.2. In MR, the discriminatory ability was very strong for RVol >64ml, RVol index >32ml/m2, and RF >41% (LVSV-AoFF) and RVol >40ml, RVol index >20ml/m2, and RF >30% (MiIF-AoFF) with a negative likelihood ratio < 0.1. In conclusion, CMR grading of chronic AR and MR should be based on modality-specific and quantification method-specific thresholds, as they differ largely from recognized guideline criteria, to assure appropriate clinical decision-making and timing of surgery.