This paper examines evaluation in reading. Evaluation refers to the processes by which readers monitor their ongoing text comprehension to assess performance and task difficulties. Monitoring is assumed to explain a large proportion of individual differences in text comprehension, in that individuals need to be aware of their objectives and difficulties in order to adjust their strategies to match task requirements. The participants were French children in grades 3 and 5. Study 1 examined their evaluation-related knowledge (i.e.. knowledge about reading objectives, sources of difficulties, comprehension awareness). Study 2 examined the children's actual evaluation behavior during a comprehension task. The results showed that few children provided elaborate verbalizable knowledge about reading tasks, goals and corresponding skills, and that in most cases the children were highly confident in their responses to text comprehension questions, regardless of the correctness of their answer. The contribution of metalinguistic awareness to literacy acquisition as well as the implications of this type of study for educational practice are discussed. (Contains 4 tables and 5 figures.)