The teaching and learning of history involves a process where students and teachers interact in order to analyse evidence, raise questions and hypotheses, synthesise facts, communicate their ideas, understand others viewpoints, consider values, reflect and engage in moral reasoning (Brophy, 1996). Further it is interaction that helps in the development of thinking in history (Coltham, 1971). This paper reports on an in-progress PhD study in which the metacognitive language demands of the history class are explored. The form that metacognitive language may take in the history classroom is not well documented, hence the purpose of this investigation. This study was conducted by examining the language use and interaction pattern in the teaching and learning of history. The data was gathered from six classes from schools in Singapore. They constituted two classes of above average students (Gifted stream), two classes of average students (Express stream) and two classes of below average (Normal stream). Audio and video recordings were made of two lessons from each of the six classes. These lessons were transcribed and analysed to ascertain which contexts were more conducive for the production of metacognitive language. It was found that a complex interrelationship of factors including type of pedagogic activity, type of teacher questions, and type of student activities determined whether or not metacognitve language was used.