Two classes of high capacity hydrogen storage materials, the metal tetrahydroborates and the metal ammines, were investigated at the atomic scale using density functional theory simulations. It was shown that simple model structures could be used to asses the stabilities of complex systems. Trends in stabilities were reproduced for known systems and the correlations were used to predict the stabilities of unknown systems. Of these, 20 tetrahydroborate systems formed stable mixtures with promising stabilities. A few mixed metal ammine systems showed promising decomposition energies but their stabilities are questionable and should be investigated further. The ab-/desorption cycles of magnesium and calcium ammines were analyzed and the faster kinetics of the magnesium ammines could be explained by a layered structure of magnesium chloride. It was found that doping calcium chloride with iodine could force it into a layered structure which is expected to improve the kinetics. Iodine doping could also be used for improving ion conduction in lithium tetrahydroborate, which is useful for batteries. Only the high temperature phase of lithium tetrahydroborate show a high ion conduction, and it was shown that doping lithium tetrahydroborate with iodine stabilizes the high temperature phase, in agreement with experiment. Finally, examples on how systematic structural studies of metal halides and hydrides can aid the design of new materials were given.