This paper considers evidence from an ESRC funded study of twenty teachers, teaching the literacy hour. In 170 hours of observation only one instance of a teacher modelling her thinking about reading or writing was recorded: and this was unplanned. It is suggested here that, although there should be opportunities for metacognitive modelling within the literacy hour, teachers find it difficult to use these opportunities. Some ideas about the importance of metacognition are reviewed and an example of metacognitive modelling in shared writing is analysed. It is argued that concern for improved performance may cause more attention to be focused on what is to be achieved rather than how.