Two studies were conducted to examine the effect of prompting self-regulation, an intervention designed to improve learning from technology-delivered instruction. In Study 1, trainees who were prompted to self-regulate gradually improved their declarative and procedural knowledge over time, relative to the other conditions, whereas test scores declined over time for trainees who were not prompted to self-regulate. In Study 2, basic performance remained stable over time and strategic performance improved over time for trainees who were prompted to self-regulate, relative to the other conditions, whereas performance declined over time for trainees who were not prompted to self-regulate. Trainees’ cognitive ability moderated the effect of the prompts on basic performance and task specific self-efficacy moderated the effect of the prompts on strategic performance. Prompting self-regulation resulted in stronger performance gains over time for trainees with higher ability or higher self-efficacy. These results demonstrate prompting self-regulation improved performance over time, relative to the other conditions, in both online, work-related training and laboratory settings. The results are consistent with theory suggesting self-regulation is a dynamic process that has a gradual effect on performance and highlight the importance of using a within-subjects design in self-regulation research.