The analysis of time-dependent simulation data is a demanding task, both in terms of computing power and time. Interactive analysis using multiple linked views has been shown to be one possible solution to this problem. However, there are two significant shortcomings when limited to a standard desktop-based setup: First, complex spatial relationships are hard to understand using only 2D projections of the data. Second, the size of today's simulation runs is too large to be handled even by powerful workstations. We describe a system for the interactive analysis of large, time-dependent data in virtual environments. Based on the techniques of multiple linked views and brushing, our approach allows the user to quickly formulate, visualize and assess hypotheses about the data. To enable an interactive exploration even in the face of multi-gigabyte data sets, we distribute the workload to a multi-processor parallel machine and a rendering client.