Is clinical pharmacology a science or only an application of science? Karl Popper suggested a method to identify science and to sort it out from other logical activities such as metaphysics, whereby the falsification criterion he proposed can apply to the theory in such a way that the theory could be refuted. The clinical pharmacologist's approach requires the build-up of a therapeutic model on the basis of two other models: the physiopathologic and the pharmacological. The three-model construct is a theory. Is it scientific in the Popperian sense? From the therapeutic model, one can predict the efficacy of a drug, and the corresponding statement is tested by a clinical trial. Whatever the original statement, it is modified into a refutable one because of the use of the statistical approach in clinical trials. Furthermore, the predicate represents a hypothesis of the model validity, which will then be confronted with 'reality' through clinical experiment. As the therapeutic model is refutable, clinical pharmacology is a science in the Popperian sense.