Traditional ray tracing is extremely slow: images take many minutes, even hours days to render. Several methods have been developed over the last 15 years to speed up ray tracing, but remain unused from either lack of awareness or apparent complexity of implementation. In this paper, we examine one such technique: uniform spatial subdivision, in which space is divided into a regular grid of voxels which can be traversed efficiently. We give a simple method for the preprocessing stage, and then examine a method of traversal not much more complicated than Bresenham's line-drawing algorithm. We conclude with some empirical evidence of the vast speed increase spatial subdivision can afford. We hope this exposition will allow a more widespread use of spatial subdivision, resulting in shorter rendering times and more fruitful experimentation.