Inherent uncertainty in geological data acquisition causes the generation of large sequences of multiple equiprobable 3D reservoir models. It is not feasible to perform computationally costly numerical flow simulations in such a vast solution space; rather, a suitable approach is to carefully select a few geological models that reasonably represent the overall variability of the reservoir ensemble. The search and selection of these representative models are critical tasks that provide the Oil and Gas industry economic ways of uncovering symbolic production forecasts. This project leverages Virtual Reality (VR) to provide engineers a tool to run geological uncertainty analysis, allowing them to perform intensely spatial tasks by using an associative 3D interaction space. This work includes an analytical framework for uncertainty assessment of geological reservoir ensembles, as well as a new VR system that makes use of such underlying framework.