The aim of the present thesis was to investigate high-school students metamemory and metacomprehension of texts. In three studies the students read texts and then made prospective as well as retrospective ratings of their own immediate and delayed performance (i.e., measured via text recall and answering performance of comprehension questions). The data have been viewed overall and for different verbal skill groups. Different types of instructions, time of test, placement of rating, types of texts and characeristics of texts have been used. The overall pattern of data suggests that the students accurately predicted and postdicted their text recall. Delayed postdiction accuracy was found, even after a long delay. The pattern for comprehension was not as straightforward, in the sense that the studies demonstrated different results regarding calibration accuracy. However, the students postcalibrated more accurately their comprehension. From a verbal skill perspective, high performing students excelled in performance but the low performing students made the most accurate ratings of memory performance. Irrespective of verbal skill, the students demonstrated study preferences for both memory and comprehension of texts. These preferences interacted with text recall but not with answering performance on the comprehension questions. The results suggest that effort is a key concept to consider in this line of research. First, the students found reading to remember a more effort requiring task than reading to comprehend. This supposedly resulted in better awareness of memory performance than comprehension of the same texts. Also, the reading instruction that empasizes learning, yielded both immediate and delayed prediction accuracy. This instruction was regarded as the most effort requiring. Second, the better the person's verbal ability, the less attention he or she requires to complete the reading task, with the best possible outcome as a result. High verbal skill reading is presumably effortless and automatized. Third, when students studied texts in their most preferred way it again resulted in best possible text recall, but reduced prediction accuracies. Taken together, metacognitive thinking seems to be most useful in the beginning of and in the development of skill.