In Germany two continuous research services are available measurement the relative readership of the editorial elements in 30 major medical magazines. One service is private, the other one is part of the syndicated audience research organisation LA-MED. Both services apply, broadly speaking, the same method: The doctors are asked to mark, while reading, every editorial matter they look at in the magazine. The copies are mailed to the research institute where they are analysed. The editing house Medical Tribune, Wiesbaden, has been using the results of this research since 1976 in order to improve the quality of their publications. The editorial staff is regularly confronted to the research findings and practical consequences are drawn. The improvements mainly concern the formulation of headlines, the choice of pictures, the order in which articles are positioned through the book and, finally, the mix of topics to be treated. It has been possible over the years to draw some general conclusions out of the experiences gathered in applying this type of research to every-day-journalism.