Right ventricular (RV) volume measurements with cardiovascular magnetic resonance (CMR) is considered the gold standard, but acquisition and analysis remain time-consuming. The aim of our study was therefore to investigate the accuracy and performance of a semi-quantitative assessment of RV function in CMR, compared to the standard quantitative approach. Seventy-five subjects with pulmonary hypertension (15), anterior myocardial infarction (15), inferior myocardial infarction (15), Brugada syndrome (15) and normal subjects (15) underwent cine CMR. RV end-systolic and end-diastolic volumes were determined to calculate RV ejection fraction (EF). Four-chamber cine images were used to measure tricuspid annular plane systolic excursion (TAPSE). RV fractional shortening (RVFS) was calculated by dividing TAPSE by the RV end-diastolic length. RV EF correlated significantly with TAPSE (r = 0.62, p < 0.01) and RVFS (r = 0.67, p < 0.01). Sensitivity to predict RV dysfunction was comparable between TAPSE and RVFS, with higher specificity for RVFS, but comparable areas under the ROC curve. Intra- and inter-observer variability of RV EF was better than TAPSE (3%/4% versus 7%/15%, respectively). For routine screening in clinical practice, TAPSE and RVFS seem reliable and easy methods to identify patients with RV dysfunction. The 3D volumetric approach is preferred to assess RV function for research purposes or to evaluate treatment response.