WHAT EVERY LIBRARIAN SHOULD KNOW ABOUT ON-LINE SEARCHING Brian B. Wilks. Ottawa, Canadian Library Association, c1982. 256pp in 3-ring binder, $35.00. ISBN 0-88802-167-4. Professional. Reviewed by Sheila Bertram, Robert E. Brundin and John G. Wright. Volume 11 Number 2. 1983 March. This handbook reflects the experiences of a Canadian librarian working in an academic library and is directed to those librarians poised for technological change in their reference skills. The much increased use of computers, especially of micro-computers in small libraries, in business, and in homes, now makes online search possible, if not yet economically practicable, for a much wider public. Four aspects are treated: the historical development of information retrieval, the basic commands for searching nine data base systems currently available to Canadian users, developing reference strategies, and the cost factors to be considered in installing such services. The author hopes this handbook will help the beginner and also serve as a ready reference for the experienced searcher. That he only partially succeeds in meeting both ends of the spectrum does not diminish the usefulness of the information provided. The "beginner" in this context is a librarian with initial training and experience in library automation, an attainment that excludes too many present practitioners; and the data bases so far available primarily serve a highly specialized clientele more likely to be found in academic and special libraries than in small public or school libraries. The differences in entering the data base systems and in using their command structures are clearly explained and would be even more useful for quick reference if they were in tabular form. The treatment of Boolean logic and the development of search strategies need more specific examples, and will benefit most those who have an instructor to answer their questions. The final section on costing pinpoints accurately the issues to be considered but might have provided more information about the range of printed publications needed to understand how to use data bases. Actual costs, actual commands, and actual systems are constantly changing and will outdate many of the specific details in a short time. Perhaps that is the reason the loose-leaf format was used, to enable purchasers to update their own copies. The work is well-organized and well-written, although it is surprising in a Canadian context that the search examples did not mention spelling variations in key word searching such as "labour" and "labor." All told, this is a publication that highlights the technological revolution inevitable for even the smallest libraries; it will best serve the librarian who is still in a learning rather than in a working environment. Sheila Bertram, Robert E. Brundin, and John G. Wright, Faculty of Library Science, University of Alberta, Edmonton, AB.