In this overview, the methods for assessing antacid activity in vitro are surveyed, and the problems of their comparison with in vivo methods of evaluation are discussed. In vitro assessment is based on two types of method: static and dynamic. The static method of titration, with end-of-titration pH values ranging between 3.0 and 1.0, has been used to quantify the number of sites capable of binding H+ ions at each end-of-titration pH, and to identify certain chemical mechanisms involved in this binding; in other words, this approach provides the pharmacological characteristics of the drugs tested. In contrast, it does not take into account physiological factors modulating antacid activity, such as gastroduodenal fluxes (including gastric emptying), drug adherence to the mucosa, and acid secretion. The dynamic method was initially based on an artificial stomach model, which has gradually been upgraded to a computer-controlled artificial stomach-duodenum model. This model overcomes certain weaknesses of the static method by simulating flux and pH conditions in the gastroduodenal tract, by taking into account interactions with the gastric mucosa and thereby reproducing the in vivo medium encountered by antacids. It is therefore capable of reflecting the characteristics of antacids, namely their effect on gastric pH and resistance to acidification, at the same time helping to identify the underlying chemicophysical mechanisms. In vivo, the antacid effect can be assessed qualitatively by means of pH-meter studies in healthy volunteers, both in baseline conditions and during secretory stimulation, and also quantitatively by methods based on intragastric titration in response to a liquid meal (IGT). pH-meter studies in baseline conditions come up against the variability of the basal pH and antacid homogenization with gastric contents, which results in a wide range of individual values. This variability is found in pH-meter studies during pentagastrin infusion and, to a lesser degree, in response to a meal. Close correlations have, however, been established between results obtained with the artificial stomach model and in healthy volunteers submitted to pH-metric or IGT studies, with several antacids. It seems that the artificial stomach method is sufficiently reproducible to make it the method of choice for investigating the antacid activity of all drugs aimed at treating acid hypersecretion disorders. In contrast, in vivo studies may be warranted for precise therapeutic indications, such as treatment of duodenal ulcer or gastro-esophageal reflux, in which the therapeutic effect is judged on the basis of an improvement in symptoms and endoscopic criteria, without the need to demonstrate the antacid effect itself.