BACKGROUND The aim of the study was to evaluate the interobserver variability and reproducibility of the renal scintigraphy reports. METHODS The study was prospective performed in 60 99mTc DMSA scintigraphies (116 kidneys), done in one centre, displayed on X ray films as analogue 300 kcounts images in 4 standard projections. Six observers from 4 different centres (group A), including three observers from one institution (group B), were asked to interpret original scans. Seven parameters of scintigraphy protocol were analysed with at least two responses: kidney size (1), uptake (2), outlines (3), scars (4), SOL (5), differential function (6) and conclusion (7). RESULTS Among 116 kidneys, concordant results in interpretation of seven points of report from 1 to 7, in the group A were found in 62%, 42.5%, 45%, 9%, 47%, 52% and 34%, and In the group B in 72%, 55%, 59%, 22%, 62%, 60% and 41% respectively. Some improvement in concordant reporting was noted in the group of 3 observers, in comparison of group of all, statistically significant in category of scar reports (p < 0.01). CONCLUSIONS Results suggest to use of standardised criteria and terminology in interpretations, to improve low interobserver reproducibility and objectivity of renal scintigraphy reports, particularly in evaluation of renal scars.