The metaverse enables learners to engage in inquiry and interact within virtual spaces, offering high levels of immersion, interactivity, and presence. However, previous studies have primarily focused on affective outcomes, and research on learners’ higher-order cognition and knowledge generation remains insufficient. This study examined the effects of metaverse-based ecosystem learning on high school students’ biological knowledge generation using functional near-infrared spectroscopy (fNIRS) to measure prefrontal cortex activation. Twenty-seven 11th-grade students at a local high school were participated, with one class assigned to the experimental group (metaverse-based learning, n = 15) and another to the control group (video-based learning, n = 12). During the pretest and posttest, students completed biological knowledge generation tasks while brain activation was recorded. The experimental group showed significantly more activation than the control group in the right dorsolateral prefrontal cortex (DLPFC), frontopolar prefrontal cortex (FP), and bilateral orbitofrontal cortex (OFC), indicating that learners effectively engaged in observation, question, hypothesis, experimental design, and hypothesis evaluation—key processes of knowledge generation—along with heightened cognitive and emotional engagement. These findings support the effectiveness of immersive inquiry-based learning and neuroscientifically informed instructional design and highlight the need to examine generalizability across diverse learning contexts.