A study on students' metacognitive strategies was carried out with over a thousand secondary and pre-university students from 12 schools. A questionnaire adapted from Biggs (1987) was administered to students at various levels (Secondary 2, Secondary 4, Pre-University 1), from different academic tracks (General, Science, Arts) and academic streams (Special, Express, and Normal). They were required to self-report on their metacognitive beliefs; their use of metacognitive strategies in mental tasks involving memory, problem solving and comprehension; and their attitudes towards the learning of various academic subjects. 20 items from the questionnaire which were related to problem solving were categorized into four stages, namely, orientation, organisation, execution and verification and data from these items were analysed.

Some findings that emerged were:

• (a) Normal stream students exhibited a lower usage of metacognitive strategies as compared to students from the Express and Special streams.

• (b) Metacognitive strategies used by Normal stream students tended to be of the “surface” type.

• (c) There was no significant difference in the frequency of usage of metacognitive strategies between students from different academic tracks.

• (d) During the problem solving process, students spent most time on evaluation of answers rather than on monitoring their understanding.

• (e) Students from different levels (Secondary 2, Secondary 4 and Pre-University) exhibited similar frequency of usage of metacognitive strategies in problem solving.

• The implications of these findings on future research and development projects as well as the teaching of metacognitive strategies are discussed in the paper.