This book provides guidelines for evaluating a variety of database types, including abstracting and indexing, directory, full-text, and page-image databases available in online and/or CD-ROM formats. The book discusses the purpose and techniques of comparing and evaluating the most important characteristics of textual databases, such as their scope, dimensions, source coverage, record content, accuracy, consistency, currency, and completeness. Evaluation models and procedures are given for widely used databases (e.g., Books in Print, Ulrich's Periodicals Directory, Library and Information Science Abstracts and their full-text version), whose subject, content, and terminology are familiar to librarians and information specialists irrespective of their specialization. Beyond traditional online and CD-ROM databases, the book illustrates the evaluation of the new World Wide Web-born information resources (e.g., Amazon, Barnes and Noble, The Electric Library, and Northern Light). Dozens of other databases are used to illustrate significant differences of purported similar coverage (e.g., PsycINFO and Mental Health Abstracts). The book includes 12 chapters that cover: database products, producers, and publishers; database content evaluation criteria; database subject scope; database dimensions; database source coverage; record content; accuracy; format and content consistency; completeness; quality of subject indexing; quality of abstracts; and cost considerations. Includes an index. (Contains 212 references and approximately 360 illustrations/figures.) (MES)