The origin and earliest history of the earth's hydrosphere, the inventory of excess volatiles defined by Rubey in 1951, can be constrained within wide but useful limits by a consideration of empirical and theoretical evidence from astrophysics and geology. Models for the evolution of the solar system from the protoplanetary nebula and for the growth of the earth to its present dimensions suggest quite strongly that the hydrosphere came into being during accretion. Its format, with H2O mostly in the oceans, CO2 mostly in sediments, and a residual atmosphere dominated by N2, CO2, and H2O, was established at a very early date and has persisted without large, destabilizing climatic excursions until the present day. Alternative accounts of early history, in which the earth either loses a massive primordial atmosphere or acquires its secondary atmosphere by gradual degassing, seem improbable on the basis of a series of circumstantial but cumulatively persuasive arguments. The difficulty of dissipating a massive atmosphere of solar composition in reasonable times, the likelihood that accretion was a highly energetic process and that it triggered early segregation of the core, and the tendency of the planet to accumulate volatiles preferentially in the later stages of accretion are examples of arguments favoring an early origin for the hydrosphere. Several geological isotope systems which can be sampled today require early separation of the atmosphere and probably the hydrosphere as a whole; these systems record radiogenic enrichment patterns in the noble gases and stable isotope fractionations which suggest an early origin of the biosphere. Certain geological indicators of atmospheric composition, and the broadly equable character of the rock record, are also consistent with a hydrosphere established in the earliest stages of history and having an initial neutral or weakly reduced composition. The climatic stability of such a hydrosphere poses crucial questions, especially since the early sun was less luminous: we find that it is not paradoxical to have an early earth with surface temperature close to that of the present day if the major hydrospheric constituent is liquid water. A near‐global ocean and a cloudy atmosphere provide thermal feedbacks which suffice to maintain climatic stability against the perils of the “deep freeze” and the “runaway greenhouse.” Large quanltities of atmospheric greenhouse gases are not required for climatic stability, although CO2 may well have been present in amounts larger than today's; large expanses of dry land are, however, difficult to accommodate in the picture which we present, but the evidence favors differentiation of the continents later in geological history than the early times which we consider. Localized glacier ice is permitted, and perhaps likely, and may provide a milieu favorable for chemical evolution.