Metagenome data sets present a qualitatively different assembly problem than traditional single-organism whole-genome shotgun (WGS) assembly. The unique aspects of such projects include the presence of a potentially large number of distinct organisms and their representation in the data set at widely different fractions. In addition, multiple closely related strains could be present, which would be difficult to assemble separately. Failure to take these issues into account can result in poor assemblies that either jumble together different strains or which fail to yield useful results. The DOE Joint Genome Institute has sequenced a number of metagenomic projects and plans to considerably increase this number in the coming year. As a result, the JGI has a need for high-throughput tools and techniques for handling metagenome projects. We present the techniques developed to handle metagenome assemblies in a high-throughput environment. This includes a streamlinedassembly wrapper, based on the JGI?s in-house WGS assembler, Jazz. It also includes the selection of sensible defaults targeted for metagenome data sets, as well as quality control automation for cleaning up the raw results. While analysis is ongoing, we will discuss preliminary assessments of the quality of the assembly results (http://fames.jgi-psf.org).