It is increasingly difficult for the scientist to keep in touch with research in his technical field because of the increasing number of publications. Scientific information systems of an electronic nature are probably the only way out of this dilemma. In order to be able to offer the best form of service, the modern documentation and information center must know the exact requirements and wishes of their potential users. In this papter the authors are concerned with the methods and results of such user analyses as they have been carried out up to now, principally in England and the USA. In spite of various insufficiencies of method, the results of these analyses have given many interesting glimpses into the standards of behavior of "consumers" of scientific literature.