Two possible scenarios in early terrestrial atmospheric evolution are examined using a one-dimensional chemistry and flow model of the atmosphere. In each case the production of oxygen results from photolysis of H2O followed by the escape of hydrogen to space. In case 1 the rate of release of reduced volcanic gases is assumed to be greater than the oxygen production rate. This leads to ground-level oxygen concentrations on the order of 10-13 PAL(present atmospheric level). In case 2 the volcanic reduced gas source is omitted, as might have been the case during an extended period of decreased tectonic activity. The oxygen concentration would then have been limited to ~4¿10-8 PAL by reaction with dissolved ferrous iron in the early oceans. The case 1 atmosphere is shown to be reducing, and the case 2 atmosphere oxidizing, based on the relative concentrations of reduced versus oxidized radical species present in the troposphere. Redbeds should have been able to form under case 2 conditions, but not under case 1. The NO produced by lighting discharges is converted primarily to HNO in case 1 and to HNO3 in case 2. In either case, fixed nitrogen was probably sufficiently scare to be a limiting nutrient for marine organisms.