Adaptive information systems promise to augment human information seeking by cooperating with users to manage the cognitive burden of the search. Highly interactive (hypertext) systems for brows- ing databases can promote natural, low-cognitive load search strategies. Still lacking, however, is a scientific basis for designing user-oriented organizations of data or user-oriented views of database content. This research seeks to test aspects of a theory of information seeking which can explain users' interactions with machines to find information. In particular, through experimental analysis of such interactions it examines users' development and application of mental models of hypertext systems employed for that purpose. The research will contribute to an emerging theory of information seeking in electronic environments, guide designers of adaptive information systems, and facilitate the development of methods and materials for human information seeking.