Purpose of the study: This study aimed to examine the effects of augmented reality–based instruction, video-based instruction, and traditional instruction on fourth-grade students’ environmental literacy, encompassing environmental knowledge, environmental attitudes, and pro-environmental behavior, within an Adiwiyata-certified elementary eco-school. Methodology: A quasi-experimental design was employed involving 81 fourth-grade students from an Adiwiyata-certified elementary eco-school in Semarang, Indonesia. Data were collected using a validated multiple-choice environmental knowledge test and Likert-scale questionnaires measuring environmental attitudes and pro-environmental behavior. Data were analyzed using multivariate analysis of covariance (MANCOVA) and analysis of covariance (ANCOVA) with IBM SPSS Statistics version 26. Main Findings: The results revealed a significant multivariate effect of instructional method on overall environmental literacy. Augmented reality–based instruction produced the highest post-test scores in environmental knowledge and environmental attitudes, followed by video-based instruction and traditional instruction. Instructional method also exerted a statistically significant but smaller effect on pro-environmental behavior. Novelty/Originality of this study: This study provides novel empirical evidence through a three-arm comparison of augmented reality, video-based, and traditional instruction within the Adiwiyata (eco school) framework. It advances environmental education research by demonstrating differentiated cognitive, affective, and behavioral effects of immersive and non-immersive instructional methods at the elementary school level.