Serious educational gameplay continues to gain traction in K-12 educational settings, as access to novel technologies increases. Increased access to computers and other devices has shifted the landscape of the classroom, increasing students' screen-time, as technology infused lessons replace other learning experiences. Amid this context, our research team created a serious educational game (SEG) designed for elementary science learners. Previous studies have linked serious gameplay to learning gains and changes in attitudes, yet the rigor of the studies often underwhelms, as the studies lack an adequate sample size, detailed student data, or quality comparison groups. These limitations are compounded in the study of science learning in elementary schools, due to structural difficulties and student variability that impede the implementation of this type of experimentation. This research attends to this knowledge gap, as our team conducted two quasi-experimental studies to explore the learning gains associated with Virtual Vet, a SEG designed to address the fundamental science concepts associated with the human body systems and the subsequent effect of diabetes and obesity related illnesses on the given body systems. In Study 1, students (n=232) were randomly assigned to play Virtual Vet or participate in guided inquiry lessons that addressed the same objectives. Our study found that students in the game condition outperformed the non-game condition. Based on this finding, we designed Study 2 to compare learning outcomes associated with two conditions: students play only Virtual Vet or students play Virtual Vet and they participate in the guided inquiry activities, which doubled the instructional time for this group. Students (n=166) were assigned to one of the conditions and we found no significant increase in learning gains for the students who participated in both activities. These results indicate that well-designed SEGs powerfully facilitate and support science learning in elementary settings.