Educational hypermedia has been heralded as providing instruction that accommodates learners' individual differences, allowing them to learn in accordance with their unique needs, desires, and preferences. Advocates of hypermedia point to learner control, multimedia capability, and parallels to theories of human memory as evidence for educational hypermedia1s efficacy. While intuitively appealing, empirical research has not confirmed (or rejected) this assumption. Nor is there compelling evidence to suggest that learners who posses different cognitive styles benefit equally from educational hypermedia environments. This study reports on an examination of cognitive style field dependence/independence and learner control of presentation mode within an educational hypermedia environment. Participants were 122 high school juniors and seniors enrolled in classes whose curricula dealt with the political process. The experimental sites were three high schools located in North Carolina, Virginia, and West Virginia. Learners classified as field dependent and field independent were asked to explore a commercial hypermedia program concerning the presidential primary election process. Participants were randomly assigned to one of two versions of the program, a control group which utilized the commercially designed multimedia presentations and a treatment group that offered the choice of single-channel presentations in addition to the multiple-channel presentations. After interacting with the program participants completed a ten item post test consisting of five recall and five problem-solving questions. Their actions within the hypermedia environment were automatically tracked by the computer. Data analysis consisted of a 3 x 2 ANOVA to examine interaction between field dependency and learner control of presentation mode as measured by the post test. Field dependency was correlated with frequency of multiple-channel selections. A secondary analysis examined main effects and interactions on the recall and problem-solving questions separately utilizing a 3 x 2 ANOVA. Data analysis revealed no correlation between field dependency and frequency of multimedia selections. The study found no significant interaction between field dependency and control (program or learner) of presentation mode as measured by the ten item post test. Separate analysis of the recall and problem-solving questions also revealed no significant interaction. / Ph. D.