The results of clinical trials are only valid if the right statistical test for the type of study performed is applied. The authors suggest a method, accessible to all physicians, that should facilitate the choice between various tests and enable the readers of clinical trial papers to detect errors in the presentation of the results. Starting from an answer to four questions (what kind of data is investigated, how many samples are studied and are they independent, and what is the purpose of the study?) the mental process is developed in the form of a flow chart, each terminal branch of which leads to the correct statistical test or tests.