This study investigated the influence of cognitive support for learning computer-based tasks using animated demonstration (AD) on instructional efficiency. Cognitive support included (1) segmentation and learner control introducing interactive devices that allow content sequencing through a navigational menu, and content pacing through stop and play buttons along with a slide bar for fast forward and rewind; (2) presentation modality and (3) a hybrid approach using a combined format of ADs with static graphics. Instructional efficiency was a combined effect of time-on-task during the learning phase and test performance. Time-on-task was included as a measurement to determine whether this would provide an alternative explanation for the benefits of learner control. Three learner-paced conditions with different types of cognitive support were developed, whereas a system-paced format, created without additional cognitive support, was used as a control condition. The learning task of this study was using graphic vector-based software Bezier tools to trace a cartoon. The results tended to suggest that the combination of different types of cognitive support (i.e. presenting the accompanying explanations as written captions to the learner-paced AD in visual-only instruction or combining learner-paced AD with static graphics in dual-modality instruction) had a positive effect, while learner pacing alone had not a significant positive effect.