The interpretation of gastrin release is confused because of variation in the technique of the radioimmuoassay of gastrin, the lack of a standard stimulus for the release of gastrin and diversity in the method used to express the results. These problems have been analysed (a) by examining the cross-reactivity of three gastrin antisera and using each of the antisera to measure basal gastrin levels in fasting normal subjects, duodenal ulcer and post-vagotomy patients; (b) by determining a satisfactory stimulus for gastrin release in normal subjects; (c) by examining the results to determine the best method of presenting the data. The different a ntisera used were found to give different levels of plasma gastrin in the same sample of plasma. This was not related to the cross reactivity of the antisera. An English breakfast was found to be the most satisfactory stimulus for the release of gastrin. The expression of the results of such a stimulus of gastrin release was affected least by assay variation when the incremental integrated gastrin response was used. It is concluded that the incremental integrated gastrin response to an English breakfast is a satisfactory method for exploring variations in gastrin release.