Attentional engagement is known to vary on a moment-to-moment basis. However, few self-report methods can effectively capture dynamic fluctuations in attentional engagement over time. In the current paper, we evaluated the utility of stimulated recall, a method wherein individuals are asked to remember their subjective states while using a mnemonic cue, for the measurement of temporal changes in attentional engagement. Across three experiments, we asked participants to watch video lectures and assessed their in-the-moment levels of attentional engagement using intermittent thought probes. Then, we used stimulated recall by cueing participants with short video clips from the lectures to retrospectively assess the levels of attentional engagement they had experienced when they first watched those clips within the video lectures. Experiment 1 tested the statistical overlap between in-the-moment and video-stimulated ratings, Experiment 2 tested the generalizability of stimulated recall across different types of videos, and Experiment 3 tested order effects by presenting the video-stimulated probe clips out of sequential order. Across all experiments, we found statistically robust correspondence between in-the-moment and video-stimulated ratings of attentional engagement, illustrating a strong convergence between these two methods of assessment. Taken together, our findings indicate that stimulated recall provides a new and practical methodological approach that can accurately capture dynamic fluctuations in subjective attentional states over time.