Self-regulation in flipped learning is important because it is positively related to academic performance. Thus, flipped learning environments must be designed to incorporate active learning approaches that enhance self-regulation. For flipped learning to be successful, both online and physical learning components are important. This project aims to investigate active learning approaches (collaborative and online interactive learning) for flipped learning to improve self-regulation among polytechnic learners in applied science, and to determine the extent to which self-regulation in flipped learning environments can improve learners’ academic performance. A mixed-methods, equivalent time-series design-based research approach was employed in this study, with the first two cycles focused on improving the intervention. In the final cycle, an evaluation of the intervention was conducted. Test scores were utilised to determine academic performance, while the Motivation Strategies for Learning Questionnaire was used to measure self-regulation. The results show that both self-regulation and academic performance improved. For both active learning approaches, dimensions such as metacognitive self-regulation, goal orientation, and task value improved significantly. Academic performance increased after the interventions, though Spearman’s rank correlation analysis revealed only weak to moderate relationships between self-regulation and academic performance, suggesting the influence of other factors, such as motivation. The study contributed practical guidelines for educators, offering evidence-based strategies to enhance self-regulation in flipped learning environments. Unlike existing literature that emphasizes personal strategies (e.g., time management), this research highlighted the role of structured pedagogical approaches.