Immersive virtual reality (IVR) technology influences students’ cognitive processes and affective experience, according to previous research findings. Yet, empirical studies that investigate the impact of IVR on science education in secondary schools are limited. This article reports a pilot study that has two objectives: First, to investigate whether learning science with IVR leads to more positive learning outcomes than that with a traditional video-viewing activity; and second, to investigate whether corrective feedback produces more positive learning outcomes than explanatory feedback in IVR-mediated contexts. A randomized, controlled pilot trial was designed during a school year at a secondary school in western China. Forty-seven participatory students were randomly assigned to a control group (learning with instructional videos, n = 17), an intervention group 1 (learning with IVR with corrective feedback, n = 15), and an intervention group 2 (learning with IVR with explanatory feedback, n = 15). We evaluated an array of student learning-related outcomes, including extraneous and germane cognitive loads, intrinsic learning motivation, academic self-efficacy, learner satisfaction, and academic achievement. The results revealed a significant difference in intrinsic learning motivation, academic self-efficacy, learner satisfaction, and academic achievement, but an insignificant difference in extraneous and germane cognitive loads when students learned with video and IVR. In terms of the type of feedback provided in IVR-mediated environments, students who were prompted with corrective feedback and explanatory feedback showed little difference in their learning outcomes. The reasons related to the findings and limitations are discussed.