Issues are discussed that are critical in determining tritium loss from helium-cooled fusion breeding blankets. These issues are: (a) applicability of present models to permeation at low tritium pressures, (b) effectiveness of oxide layers in reducing permeation, (c) effectiveness of hydrogen addition as a means to lower tritium permeation, and (d) effectiveness of conversion to tritiated water and subsequent trapping as a means to reduce permeation. The paper discusses theoretical models applicable to these issues, and presents results of experiments in two areas: permeation of mixtures of hydrogen isotopes and conversion to tritiated water.