Multimedia principles such as segmentation, learner control, and modality have been proposed to ensure animated demonstrations (AD) effectiveness. However, several studies that apply multimedia principles in different learning contexts with different learners have found boundary conditions for several effects. This study thus aimed to investigate different levels of learner control, segmenting, and modality effects on learning and cognitive load when AD was used as the before-class instructions in the flipped classroom, in which the boundary conditions include self-pacing and sufficient learning time. Two experiments (low control vs. high control) were conducted with a 2 × 2 (segmentation: nonsegmented vs. segmented; modality: narrated vs. captioned) between-subjects ANOVA design. A supplementary analysis across the two experiments was further conducted to distinguish between the main effects of learner control and segmentation. The findings of this study have important implications for the design of before-class instructions to enhance student engagement and understanding.