Imaging with 99mTc-IDA derivatives is a recognised technique for assessing duodenogastric reflux. A study of 150 consecutive scans has revealed a number of limitations in its use. Anatomical definition of the stomach on static imaging and computed data analysis of reflux, were often complicated by overlap by the left lobe of the liver and the duodenal-jejunal flexure. Delineation of the antrum was poor, and as it is so close to the duodenal cap and hepato-biliary activity, transient small volumes of reflux may not have been detectable. Quantification of massive reflux was easily reproducible but computer analysis of small volumes of transient reflux was often uninterpretable. A proposed clinical grading of reflux takes into account the intensity, extent and duration of reflux, and can be matched with other clinico-pathological features.