Purpose: The purpose of this research is to investigate student engagement in guided web-based learning systems. It looks into students' engagement and their behavioral patterns in two types of guided learning systems (i.e. a fully- and a partially-guided). The research also aims to demonstrate how the engagement evolves from the beginning towards the end of the interactions; which enables analysis to be performed on the quality of engagement. Design/Methodology/Approach: An experimental study was conducted on 41 students from a public university in Malaysia using two web-based systems as the main learning tools. The students' engagement data were captured three times during the interactions and once at the end of the experimental study using student self-report. Findings: The main outcome of this study suggests that student engagement was changing over time either in positive or negative patterns. The directions of change in both types of guided learning were mainly influenced by the students' background of knowledge. Practical Implications:This study demonstrates that student engagement is dynamic. Therefore, progressive assessment is a practical approach to obtain the engagement data which can be used to regulate and improve student engagement in web-based systems. As a result, an adaptive and intelligent web-based learning environment can be created. Originality/Value: This research proposes a new approach to improve students' engagement in web-based instruction, that is, through a progressive assessment of their current experience. (Contains 5 tables, 5 figures, and 2 notes.)