Borosilicate glass is the principal solid matrix for immobilizing 99% of the highly radioactive, heat-generating nuclides extracted by reprocessing spent nuclear fuel. Production of the glass has begun in several countries, yet no final disposal site is available anywhere in the world. This is due partly to political issues and partly to the difficulties of credibly demonstrating that nuclear waste can be safely isolated in deep underground repositories for hundreds of generations. The release of hazardous quantities of radionuclides from a repository is prevented by a multiple barrier containment system, including a central engineered system consisting of the canistered glass, an overpack, and backfill materials. If the glass could retain all radionuclides upon contact with groundwater, it would not be necessary to demonstrate that geological isolation is safe. However, the glass corrodes slowly in water and humid air, and inevitably, certain quantities of radionuclides are mobilized. The glass is not inherently corrosion-resistant, but rather depends on the waste package and on surrounding geochemical and hydrological constraints. The difficulty is predicting the release/retention of radionuclides for long time periods while considering the interactions with other engineered and natural barriers.