On the basis of a model for communication, a series of requirements are established for the content and form of written information given in connection with clinical trials. Fourteen texts employed in an oncological department are reviewed. The clarity as expressed by the readability index (lixtal) is found to be considerably less than desired. A series of examples are presented which either make reading difficult or are such that the patient does not receive adequate and relevant information. In particular, information about alternatives to participation in the trial and the possibilities for feedback for the patient are not mentioned. Finally, proposals are made for how the doctor can ensure that the written information is adequate and readable.