Even though learning refers to both a process and a product, the former tends to be overlooked in educational virtual reality (VR) research. This study examines the process of learning with VR technology using the Cognitive Affective Model of Immersive Learning (CAMIL) as its framework. The CAMIL theorizes that two technological features of VR, interactivity and immersion, influence a number of cognitive and affective variables that may facilitate or hinder learning. In addition, VR studies often involve media comparisons that make it difficult to disentangle the relative effects of technological features on learning. Therefore, this study also aims to provide insights concerning the unique and combined effects of interactivity and immersion on the cognitive and affective variables specified by CAMIL. We employed a 2 × 2 between-subjects design (N = 153) and manipulated the degree of interactivity and immersion during a virtual lesson on the topic of viral diseases. Analyses of variance (ANOVAs) were used to examine the effects of interactivity and immersion on our variables of interest, and structural equation modeling (SEM) was used to assess the process of learning as predicted by the CAMIL. The results indicated that the process of learning involves situational interest and embodied learning. Main effects of interactivity and/or immersion on cognitive load, situational interest, and physical presence are also reported in addition to interaction effects between immersion and interactivity on agency and embodied learning. The findings provide evidence for the CAMIL and suggest important additions to the model. These findings can be used to provide a better understanding of the process of learning in immersive VR and guide future immersive learning research.