This study investigated how the frequency and level of assistance provided to students interacted with prior knowledge to affect learning in the Voyage to Galapagos (VTG) science inquiry-learning environment. VTG provides students with the opportunity to do simulated science field work in Galapagos as they investigate the key biology principles of variation, biological function, and natural selection. Thirteen teachers used the VTG module during their Natural Selection and Evolution curriculum unit. Students (N = 1728) were randomly assigned to one of four assistance conditions (Minimal-, Medium-, Medium–High, or High-Assistance). We predicted we would find an “Expertise Reversal Effect” (Kalyuga et al. in Edu Psychol Rev 194:509–539, 2007), whereby students with little prior knowledge benefit from assistance and students with higher prior knowledge benefit from minimal assistance. However, initial analyses revealed no interaction between prior knowledge and condition on student learning. To further explore results, we grouped students into 5 clusters based on student behaviors recorded during the use of VTG. The effect of assistance conditions within these clusters showed that, in two of the five clusters, results were consistent with the Expertise Reversal Effect. However, in two other clusters, the effect was reversed such that students with low prior knowledge benefited from lower amounts of assistance and vice versa. Though this study has not identified which specific characteristics determine optimal assistance levels, it suggests that prior knowledge is not sufficient for determining when students will differentially benefit from assistance. We propose that other factors such as self-regulated learning should be investigated in future research.