During the two-year Relevance Assessment Project, 15 studies were designed and carried out with over 500 subjects used as relevance judges. The subjects were librarians, information specialists, library science students and faculty, and graduate and upper division students in psychology. Materials for judging were selected for subjects according to their interests and backgrounds. Results of the many experiments conducted indicate that relevance judgments can be influenced by many factors including (1) the skills and attitudes of the judges used, (2) the documents and document sets used, (3) the particular information requirement statements, (4) the instructions and setting in which the judgments take place, (5) the concepts and definitions of relevance employed in the judgments, and (6) the type of rating scale or other instrument used to express the judgments. Findings indicate that relevance scores must be used with caution for system evaluation. A list of articles and reports on the project is included. (CC)