We explore the usefulness of the in situ cosmogenic 3H, 14C, and 10Be produced by spallation of oxygen nuclei in ice, as tracers to determine net accurmulation/ablation rates of ice sheets. The application of the in situ 3H and 10Be is severely constrained because at deposition, ice contains appreciable amounts of these isotopes from the atmosphere. The case is much more favorable for 14C, which is not carried with wet precipitations; atmospheric 14C gets mechanically trapped in the ice during deposition. We point out that cosmogenic 14C would probably exist as 14CO in ice. This seems to be supported by the published results of Fireman and Norris. Considering their inherent amounts in the ice and the expected in situ production rates, we discuss conditions under which these isotopes can be used to study net accumulation and ablation rates. We also discuss available data on 14C and 10Be on polar ice from accumulation and ablation zones. We conclude that 3H and 14C should find wide applications in studying ice dynamics and 10Be in very special circumstances. ¿American Geophysical Union 1987