It is the purpose of this study to evaluate the efficacy of alternative diagnostic strategies in gastroenterological out-patients. Photocopies were made of the history and physical examination recorded at the first visit of 146 consecutive patients. In each case a senior gastroenterologist was asked to state the most likely diagnosis, indicating his or her confidence in the diagnostic bid, and to state which investigations he or she wanted done. The diagnostic bids were then compared with the actual diagnoses made at the Out-Patient Department. The chief result was that the diagnostic bid agreed with the actual diagnosis in 81% of the 146 cases. Agreement was more frequent in cases where the gastroenterologist was confident in the diagnostic bid than in those where the prediction was considered less certain. The high proportion of correct diagnoses suggests the adoption of a single-target strategy in most gastroenterological out-patients, i.e. a strategy aimed at confirming or excluding only that diagnosis which is considered most likely. A multi-target strategy, taking into account several diagnostic possibilities, may be reserved for that minority of cases where the primary diagnostic bid proves incorrect. It is suggested that gastroenterological out-patients may be subjected to fewer investigations if these recommendations are accepted. Only 26% of the items of information recorded in the notes were used for the diagnostic prediction, and, possibly, hospital records contain too much irrelevant information.