The radiological hazard resulting from an exposure to either tritium oxide or tritium gas is discussed and the factors contributing to the hazard are presented. From the discussion it appears that an exposure to tritium oxide vapor is 10/sup 4/ to 10/sup 5/ times more hazardous than exposure to tritium gas. Present and future sources of tritium are briefly considered and indicate that most of the tritium has been and is being released as tritium oxide. The likelihood of gaseous releases, however, is expected to increase in the future, calling to task the present general release assumption that 100% of all tritium released is as oxide. Accurate evaluation of the hazards from a gaseous release will require a knowledge of the conversion rate of tritium gas to tritium oxide. An experiment for determining the conversion rate of tritium gas to tritium oxide is presented along with some preliminary data. The conversion rates obtained for low initial concentrations (10/sup -4/ to 10/sup -1/ mCi/ml) indicate the conversion may proceed more rapidly than would be expected from an extrapolation of previous data taken at higher concentrations (10/sup -1/ to 10/sup 2/ mCi/ml).