To evaluate the diagnostic potential of magnetic resonance imaging (MRI) to diagnose and evaluate tricuspid incompetence (TI), right ventricular angiography (RVA) and MRI were compared in 51 patients. For angiographic semi-quantification a 4-grade modified Sellers classification was used. Several MRI-parameters (jet volume, jet area, number of slices with visible jet, duration of the regurgitation and volume of the right atrium) were examined concerning their validity to evaluate TI and compared to the angiographic data. Limits for MRI jet areas and volumes for each angiographic group were defined to obtain the best agreement with angiographic results. Thirty of 34 patients with angiographically visible TI (sensitivity = 88%) were diagnosed correctly with MRI. The 4 remaining patients also showed a small regurgitant jet which was, however, interpreted as physiologic. One of the 17 patients without angiographic TI was diagnosed as mild TI by MRI (specificity = 94%). MRI classification using jet volume (46/51 = 90% correct classifications) was not significantly better than the easier and faster MRI classification using jet area (43/51 = 84%). The other parameters (number of slices, duration of regurgitation and volume of the right atrium) showed considerable overlap between angiographic groups and did therefore not allow a classification into 4 grades. However, simple classification into hemodynamic relevant and irrelevant TI was possible. Thus MRI is able to diagnose TI with good sensitivity and specificity. Classifications based on jet volume and jet area showed good agreement with angiographic results. However, both parameters depend on the magnet and the pulse sequence used and must therefore be specifically determined for each imaging routine.