To determine the optimal plane of two-dimensional velocity-encoding cine (VENC) magnetic resonance (MR) imaging at the tricuspid annulus for quantification of tricuspid regurgitation (TR) and to assess the reproducibility of VENC tricuspid flow measurements. We conducted a retrospective review of MR images of 25 consecutive patients (M:F = 8:17; mean age, 58.5 ± 10.5 years) with TR. VENC was performed twice orthogonal to the tricuspid annulus plane during the end-diastolic (ED) and end-systolic (ES) phases. The TR fraction was quantified at each plane as retrograde flow/antegrade flow and additionally as retrograde flow of the ED plane/antegrade flow of the ES plane (combined plane method). The conventional method to determine the TR amount [right ventricular stroke volume (RVSV)—pulmonary antegrade flow] and TR fraction (TR amount/RVSV) was used as the reference standard. There were no differences between TR amount and retrograde flow of the ED plane (65.3 ± 43.4 vs. 70.5 ± 36.1 ml, P = 0.361) between the RVSV and the antegrade flow of the ES phase (124.2 ± 46.1 vs. 128.0 ± 45.0 ml, P = 0.612) or in TR fraction between the conventional and combined plane methods (48.8 ± 19.2 vs. 56.3 ± 24.3 %, P = 0.08). The retrograde flow of the ED phase was best correlated with TR amount [intraclass correlation coefficient (ICC) = 0.859] and antegrade flow of ES with RVSV (ICC = 0.808). The TR fraction of the combined plane method was best correlated with the conventional method (ICC = 0.694). Interobserver agreement of VENC flow measurements was excellent (ICC, 0.939–0.993). The optimal method for quantification of TR using tricuspid annular VENC was the combined plane method, which divides the retrograde flow of the ED plane by the antegrade flow of the ES plane. Tricuspid flow measurements using VENC showed excellent reproducibility.