Studying the human bronchi in vitro, and therefore sheltered from the toxicity problems inherent in human experiments, makes it possible to conduct a monofactorial analysis, disregarding the perturbations engendered by reflex phenomena, hemodynamic changes, etc. Analysing the effects of mediators on tissues may be less simple that it looks, due to the multiplicity of the cell types that are present. For example, in studying the effects of bradykinin we have shown that bradykinin is a potent contractile agent of small-diameter isolated bronchi, whereas it has no significant contractile effect on larger bronchi. The bradykinin-induced contraction results from a contractile component due to stimulation of the TP receptor, and of a relaxant component due to relaxant prostanoids. The two components of the bradykinin effects are produced by stimulation of B-2 receptors. In vitro stimulation of bronchi by LPS or interleukin-1 beta permits us to obtain hyperreactivity to bradykinin due to induction of thromboxane synthetase or isomerase rather than to induction of B-2 receptors or cyclooxygenase. Involvement of the nervous system may persist in the in vitro bronchial model, and indeed we have shown, for example, that pentamidine, well known for its tussigenic effect, is an indirect parasympathomimetic compound. Thus, study of the isolated bronchus permits an approach to the mechanisms of action of medicinal drugs. Despite the simplification provided compared to the in vivo study, analysis of bronchoreactivity on the isolated bronchus must take into account numerous parameters which interfere with the proper effects of the substances.