Rendering participating media is important for a number of domains, ranging from commercial applications (entertainment, virtual reality) to simulation systems (driving, flying and space simulators) and safety analyses (driving conditions, sign visibility). This article surveys global illumination algorithms for environments including participating media. It reviews both appearance-based and physically-based media methods, including the single scattering and the more general multiple-scattering techniques. The objective of the survey is the characterization of all these methods: Identification of their base techniques, assumptions, limitations and range of utilization. We conclude with some reflections about the suitability of the methods depending on the specific application involved, and possible future research lines.