<p>The use of educational technologies to support students’ scientific literacy has increasingly used in recent years. However, a lack of distinction between active and passive digital use has led to uncertainty about which and how digital use patterns contribute to students’ scientific literacy development. To address these gaps, this study employed a two-level modeling approach using data from 15-year-old U.S. students in the Programme for International Student Assessment (PISA) 2022. We identified patterns of active and passive digital use, along with student engagement and digital self-efficacy, and examined their associations with three subscales of scientific literacy. Structural equation modeling was then used to assess both direct and indirect associations of these factors on scientific literacy subscales. Results indicate that both passive and active digital use patterns were indirectly associated with all three scientific literacy subscales through engagement and digital self-efficacy. However, active digital use did not uniformly benefit learning; its effects depended on whether it supported meaningful engagement and appropriately calibrated self-efficacy. Furthermore, the ability to explain phenomena scientifically emerged as a foundational pathway facilitating subsequent development in other scientific literacy dimensions. These findings highlight that the educational value of digital technology lies in its capacity to foster motivational–cognitive processes, offering a nuanced understanding of digital use quality and providing actionable insights for improving scientific literacy with digital technology support.</p>