Utilizing a framework for systematic replication coupled with fidelity of implementation, we conducted a distal conceptual replication of a published intervention study that was originally completed with sixth and ninth-grade participants to better understand the critical features and transferability to preservice elementary teachers. The intervention involved learning about cardiac structure and function in virtual reality (VR) with 3-D representations and haptic-enabled feedback. We intentionally manipulated the setting, duration, and arrangement of participants, but used the same methodology. Results support learning about heart anatomy, function, and blood flow (p < 0.01), with an increase of 11.6% and a large effect size (d = 1.02). A similar change was documented for sixth-grade students (13.3%), but was much less than the 26.6% change for ninth-grade students, suggesting a conditional effect for prior knowledge since the content is a seventh-grade standard. Completion and accuracy of the process prompts during the intervention appear strongly related to participant outcomes and misconceptions are consistent with those reported previously. The results do not support claims about the education level and setting as important structural features, but dosage does have an effect and warrants further study. To fully realize the potential of VR for science education, effective models are needed and this study is a step in that direction.