A significant fraction of the tritium inventory of a fusion plant will be in the elemental form HT. A simple model is proposed to calculate early doses following an HT release. The dose is not dominated by the primary HT plume but by deposition of HT into the soil, subsequent oxidation to HTO by microorganisms and the following reemission of HTO. The difficulty of calculating HTO concentrations from a large area source is solved by defining a reemission velocity. All data available from the large scale release experiments in France (1986) and Canada (1987) are used to fit this parameter. With typical worst case conditions one gets an early dose of 0.04 Sv/kg-T as HT at 1000 m distance from the source, building wake effects included. This model can also be used to calculate HTO-release doses and predicts 0.6 Sv/kg-T as HTO for the same worst case condition. About 20 % of this dose is caused by reemission of HTO deposited into the soil. The accuracy of the model is estimated to be a factor of 2 – 2.5 up and down.