This study investigates the complex transition from cognitive potential to academic achievement by examining the multidimensional constructs of motivation and affectivity in high school STEM education. As STEM disciplines demand high levels of abstract reasoning and persistent problem-solving, understanding the psychological determinants of success becomes imperative for optimizing pedagogical outcomes. Drawing on the control–value theory of achievement emotions, this research explores how motivational beliefs and emotional experiences synergistically influence the adoption of self-regulated learning strategies. Grounded in the control–-value theory of achievement emotions, this study posits that students' subjective appraisals of control and value significantly shape their affective experiences in the classroom. The methodology involved a quantitative analysis of a sample of 150 high school students. Data collection utilized four psychometrically validated instruments adapted for the STEM context: the Motivated Strategies for Learning Questionnaire (MSLQ) to assess self-regulated learning; Goal Orientation (GO) scales to distinguish between mastery and performance approach; Implicit Theories of Intelligence (IMPL) to evaluate mindset prevalence; and the Achievement Emotions Questionnaire (AEQ) to measure discrete academic emotions. The analysis aims to demonstrate how a mastery goal orientation and a growth mindset function as positive predictors of activating emotions, such as enjoyment and hope, while concurrently mitigating deactivating states like boredom and anxiety. Furthermore, this study explores the role of self-regulated learning and motivation as critical mediators in the relationship between emotional states and academic achievement. The findings underscore the necessity for an integrated STEM instructional design that moves beyond technical content to support the development of emotional resilience and autonomous learning strategies, providing educators with targeted interventions to optimize student performance.