The discovery of high concentrations of water-ice just below the Martian surface polar areas made by Mars Odyssey has strengthened the debate about the search for life on Mars. Generally it is believed that life on Earth emerged in liquid water from the processing of organic molecules. Thus, the possible origin of life on early Mars should have been related to the evolution of the planetary water inventory, consequently it is important to know the amount of water-ice stored below the planetary surface. The search and mapping of the present subsurface water and ice reservoirs will be carried out experimentally by Mars Express with its Mars Advanced Radar for Subsurface and Ionospheric Sounding (MARSIS) ground-penetrating radar in the near future. We estimate the present and past water-ice reservoirs, which are and were in exchange with the atmosphere by using the observed D/H ratio in the atmospheric water vapour, measured D/H ratios in Martian SNC meteorites and D/H isotope ratios based on a study by Lunine et al. (2003) regarding asteroid and cometary water delivery to early Mars. Using the results of this study with initial D/H ratios of about 1.2–1.6 times the terrestrial sea water (TSW) ratio and the assumption that these ratios were not fractionated by XUV-driven hydrodynamic escape due to a more active young Sun prior to 3.5 Ga, one finds a present water-ice reservoir, which can exchange with the Martian atmosphere, equivalent to a global ocean layer with a thickness of about 3.3–15 m. By assuming that hydrodynamic escape fractionated the D/H ratio to a value that is stored in the old Martian SNC meteorites with a measured average enrichment of about 2.3 times the TSW ratio we estimate a present water-ice reservoir equivalent to a global layer with a thickness of about 11–27 m. From the obtained range of the estimated present water-ice deposit, we estimate a water-ice reservoir exchangeable with the atmosphere on Mars 3.5 Ga equivalent to a global ocean with a thickness of between 17 and 61 m. All the estimated reservoirs depend on the escape of water from Mars since 3.5 Ga equivalent to a global ocean with a thickness of about 14 m (minimum) to 34 m (maximum). The main uncertainties in the estimate of the minimal and maximal water-ice reservoir is related to the present uncertainties in the efficiency of atmospheric escape rates triggered by plasma instabilities and momentum transfer effects between the solar wind and the ionosphere. However, these uncertainties will be reduced in the near future, since both loss processes will be studied in detail by the Automatic Space Plasma Experiment with a Rotating Analyzer (ASPERA-3) on-board Mars Express. The obtained results combined with the discovery of the present water-ice subsurface reservoirs by the MARSIS radar and isotope studies as presented in this work, will also give us an idea of how enriched the atmosphere was in D compared with H after the heavy bombardment corresponding to a better understanding of the efficiency of the hydrodynamic escape process due to the young Sun.