This study employed a 2x2 (Generative Activities by Navigational Activities) post-test only research design in examining learning while using a hypertext system in a linear, navigational, linear generative, or navigational generative format. Participants were 43 undergraduate students. In the first experiment, participants were assigned to one of the four conditions and expected to learn the information to solve a posed problem, while students in the second experiment learned the information to pass a reading comprehension test. Results from the first experiment revealed that students in the navigational condition outperformed those in the linear condition with respect to their problem solving score. While there was no significant difference between generative and non-generative conditions, the study found that there was an interaction indicating that generative activity may inhibit navigational hypertext systems in problem solving. The second experiment revealed significant differences between generative and non-generative conditions. While no significant difference between linear and navigational conditions with respect to reading comprehension scores was found, learners' reading comprehension raw score and dwell time revealed that learners assigned the linear conditions outperformed those assigned navigational conditions; this effect was canceled by dwell time in the final analysis. In both experiments the computerized text used was identical. (Contains 44 references.) (MES)