Background: The complex right ventricular (RV) geometry and load dependency of indices used to quantify its function make RV assessments very challenging. Differences between right and left ventricular (LV) geometry, wall thickness, myofiber architecture, blood supply and load dependency of myocardial function, limit the use of data obtained from LV studies for RV assessment. The reliability of criteria used for weaning decisions in patients with a RV assist device (RVAD) is barely known. We assessed this issue. Methods: In 36 patients who were weaned from mechanical RV support since 2003 we analyzed echocardiographic data on RV size, geometry and function plus the data on pulmonary hemodynamics recorded during “off-pump” trials. The final goal was to identify the parameters with the highest predictive value for post-weaning stability of RV function. Wall motion velocity was measured by pulsed-wave tissue Doppler. Results: Of 36 evaluated patients, 19 were weaned from “short term” bridge to recovery designed RVADs and 17 from bridge to transplant designed biventricular assist devices (BVADs). Post-weaning right heart failure (RHF) recurrence occurred in 6 (16.7%) patients. Of these 6 recurrences, 4 occurred after BVAD removal and 2 after the removal of short term designed RVADs. To date 25 patients have already reached between 1 and 7 years of post-weaning RV stability. During final off-pump trials performed before RVAD removal all 36 patients had RV ejection fraction (RVEF) > 45%, stable central venous pressure at values ≤ 12 mm Hg and ≤ grade 2 tricuspid regurgitation (TR). The mean pulmonary arterial pressure was between 18 and 30 mmHg. Off pump RV long/short axis ratio (L/S) < 0.55 and the tricuspid annulus peak systolic velocity (TAPSm) > 8 cm/s showed the highest predictive value for > 1 year stable RV function after weaning (89.5% and 94.4%, respectively). In patients without elevated pulmonary vascular resistance both TAPSm ≤ 8 cm/s and L/S ≥ 0.55 appeared to be risk factors for early recurrence of RHF. Conclusion: RV geometry and the velocity of RV systolic wall motion during off pump trials in patients with RVEF ≥ 45% and without relevant TR are highly predictive for a stable RV function after RVAD removal.