In order to evaluate the significance of gated equilibrium radionuclide ventriculography (RNV) for non invasive quantification of valvular regurgitation and follow up, various approaches were tested concerning accuracy and reproducibility. By using in vitro labeling of red blood cells and extending the acquisition time, a clear reduction of dispersion was obtained in patients without valvular insufficiency. Quantification of regurgitation directly from functional images (ventricular amplitude or stroke volume image) was clearly superior compared to the variable region of interest method. Employing functional images, reproducibility between two observers and between two independent measurements was excellent. Correlation to regurgitation values determined by cardiac catheterization was only moderate with all RNV approaches tested. RNV is limited in the absolute quantification of valvular regurgitation due to the variable overlap of right atrium and right ventricle. However, because of its high reproducibility, RNV is a non invasive technique suitable for intraindividual follow up of patients with valvular insufficiency.