The aim of the thesis was to analyze the structure of the self-regulation competence when learning with expository texts, which is represented by several competencies and their intercorrelations. <br><br> Within the last decades, various models on self-regulated learning have been published. On the one hand, psychological research proposed process models which describe different phases of self-regulated learning (e.g. Winne & Hadwin, 1998; Zimmerman, 2000). Within these phases, phase-specific metacognitive demands have to be met for a high learning outcome. However, process models do not explicitly specify the competencies needed to meet these metacognitive demands. On the other hand, researchers have proposed component models which describe students’ competencies that foster self-regulated learning (e.g., Boekaerts, 1997). Even though component models define competencies important for becoming a successful learner, they do not explicitly specify in which phase of the learning process the different competencies are needed. Against this background, an integrated model was generated identifying ten competencies necessary to meet the phase-specific demands within the different phases of the learning process. In this model, competencies were defined as cognitive dispositions, which are needed in a certain context, for instance, when learning with expository texts. To be able to assess all identified competencies, new methods were developed which meet the demands for assessing cognitive dispositions: independence of a specific situation and focus on the quality of operations. In the first analysis of the empirical part of the thesis, all methods turned out to fit the classical test criteria. In the main study of the thesis two analyses were carried out. First, it was analyzed whether the identified competencies were predictive of learning outcome in an independent learning situation. Second, the structure of the self-regulation competence was analyzed. The design of the main study was twofold. On the first day, two different expository texts were administered (topics from chemistry or physics). Within 15 minutes, the students were asked to learn as much as possible while reading the challenging text. Before and after reading the text, content-valid achievement tests were used to assess the knowledge of the students. On a second day, the competencies were assessed using the newly developed tests. If students had worked on a chemistry text the first day, the test to assess the competencies on the second day dealt with topics from physics and vice versa. To get a first impression of how predictive the competencies are for learning outcome in an independent learning situation, the power of the competencies to predict learning outcome was calculated for the well-known competencies “application of learning strategies” and “activation of strategy knowledge” using a multiple-regression analysis. Both competencies turned out to be good predictors of learning outcome. Furthermore, models on learning strategies could be expanded using the identified competencies. Based on this, the predictability of all competencies for learning outcome in an independent learning situation was analyzed using a multiple-regression analysis. Directly compared, the competencies “application of learning strategies”, “activation of strategy knowledge” and “evaluation of the knowledge after reading an expository text” were the strongest predictors of learning outcome. Following this analysis, the structure of the self-regulation competence was analyzed. Overall, the intercorrelations between the competencies were found to be rather low. However, using an explorative factor analysis, two higher level factors could be identified dealing with the central prerequisites for learning. The first factor represented the ability to judge one’s own knowledge-base. The second factor comprised all competencies which are needed to judge the given learning conditions. Thus, the theoretical assumption of internal and external prerequisites for learning (see Winne & Hadwin, 1998; Zimmerman, 2000) could be supported assessing competencies in terms of cognitive dispositions.