The inconsistency task is used to measure metacognitive monitoring in text comprehension via online (reading time) and offline measures (number of detected inconsistencies). Few studies have examined stability in task performance and interindividual differences. We addressed these issues with adolescents (N = 341) in Grades 6/7 (Mage = 12.02 years) and 8/9 (Mage = 14.07 years) at two measurement points and with expository texts. Performance in the offline and online measures were quite stable over one year with lower stability in the online measure. The two measures were only weakly associated. Only the offline measure was related to reading comprehension skills. Older (vs. younger) students performed better in the offline measure. The effect of grade level on the offline measure was mediated by intelligence and working memory capacity. Thus, the offline measure seems to be a better indicator of individual differences in comprehension monitoring in secondary school.