Purpose: The purpose of this paper is to examine the effectiveness of problem-based learning (PBL) in higher education based on a large sample of first-year undergraduates from two programmes at a Hong Kong University (n=132). One programme uses an entirely problem-based approach to learning, whilst the other uses traditional methods. Design/methodology/approach: Using the Learning and Study Strategies Inventory (LASSI) as a measure of metacognition, differences in metacognitive development are explored between each group of students between the beginning and end of their first 15 months in each programme. Findings: Despite significantly weaker entry scores on the LASSI, the mean final scores, taken after 15 months and three semesters of study in the different curriculum environments demonstrate dramatic improvements in metacognition for the PBL group. In addition, analysis of student learning experience measured at the end of the programme revealed that the PBL group reported significantly higher scores in their overall course satisfaction and generic skills development. Practical implications: The paper argues that, in addition to the formal learning context, everyday challenges emerging from the additional new social contexts provided by problem-based curricula provide fertile environments for the development of metacognition and enhancement of the learning experience. The implications of PBL environments on fostering constructivist learning and enhancing student experience are discussed. Originality/value: This research is original in its use of the LASSI inventory as a pre- and post-measure of metacognitive development in undergraduates. This is an online questionnaire administered to two groups of students following similar programmes except one is problem based and the other more traditional, and the results are strikingly significant. (Contains 2 tables and 2 figures.)