A meal intended for use in gastric emptying studies must be highly reproducible, must provide a normal physiological stimulus in terms of bulk, calorie content and composition and must employ stable radiotracers which accurately reflect in their biodistribution, the fate of the two phases. This is particularly important in a field, such as gastric emptying, where so many variables may influence the results. A conventional pancake and orange juice were chosen as suitable vehicles for the solid and liquid phases. 111In-labelled resin beads were used as the solid-phase marker and a variety of 99Tcm-labelled radiopharmaceuticals including pertechnetate, DTPA and colloid forms were investigated as liquid-phase markers. Prior to administration to patients, the stability of the phases and their interactions in vitro were investigated. The use of 99Tcm-DTPA resulted in a loss of 111In from solid to liquid phase. All non-colloidal markers exhibited a tendency for adsorption onto solid phase. Colloidal markers including rhenium and antimony sulphide colloids showed the truest delineation of the liquid phase.