When studying, the reliability of metacognitive monitoring, the predicted achievements at test, is strongly associated with the quality of study regulation and with ultimate performance at test. Previous studies that compared learning texts on screen to learning from printed texts found that screen learners performed worse and were overconfident about their success. The present research examined two methods for overcoming screen inferiority in these respects. Gaining experience with the study-test task with six different texts allowed improvement. Writing keywords after a delay from learning already eliminated screen inferiority from the first studied texts. In both methods, predictions of performance did not reflect changes in test scores. The two methods clearly affected screen and paper learners differently. This study outlines directions for overcoming screen inferiority, but also calls attention to the effects of context on cognitive and metacognitive processes, beyond the mere interaction between the person and the task content.