There is an increasing number of datasets with many participants, variables, or both, in education and other fields that often deal with large, multilevel data structures. Once initial confirmatory hypotheses are exhausted, it can be difficult to determine how best to explore the dataset to discover hidden relationships that could help to inform future research. The purpose of this study is to examine the feasibility of applying Random Forests--a non-pragmatic data mining method that creates ensembles of simple decision trees--to efficiently explore large, multilevel datasets commonly found in educational research. This method has the potential to increase the ability for researchers to perform efficient exploratory data analysis without the common pitfalls and methodological challenges. It is important to note that because this method is non-parametric, there is no need to explicitly model random effects to account for nesting (as there are no standard errors). However, the algorithm does need to be slightly altered in its splitting and re-sampling procedure to ensure error estimates are accurate. This method is both a feasible and relatively easy-to-understand statistical tool for applied researchers to effectively explore their data to help uncover potential hidden relationships and identify variables that might have been overlooked in the confirmatory hypothesis testing phase. Tables and figures are appended.