Isosurface rendering is a technique for extracting and visualizing surfaces within a 3D volume. It is a widely used visualization method in many application areas. In this paper, we describe a component-based implementation of isosurface rendering for visualizing very large datasets in a distributed, heterogeneous environment. We use DataCutter, a component framework that supports subsetting and user-defined processing of large multi-dimensional datasets in a distributed environment. We present experimental results on a heterogeneous collection of multiprocessor machines. (Cross-referenced as UMIACS-TR-2001-34)