The Greuling-Goertzel method is applied to calculation of the slowing down of neutrons in deuterium, and the results compared with the Selengut-Goertzel method, in which the deuterium slowing-down is treated by age theory. It is shown how existing codes for calculating slowing down in hydrogen can be modified in a simple manner to incorporate this treatment of deuterium. Numerical results show excellent agreement between measured and calculated ages, and indicate that a continuous slowing-down model for deuterium is inappropriate. This is in qualitative agreement with the experiments performed by Wade, and in disagreement with Olcott's work. However, it is shown that an age kernel with an age to indium of 100 cm2 may be used to compute the fast leakage from heavy-water systems over a wide range of buckling. The situation concerning agreement with critical experiments remains to be clarified because of large uncertainties in other criticality factors.