Current theory of stratospheric photochemistry indicates that the photochemical lifetime of hydrogen peroxide (H2O2) is short enough in the middle stratosphere (1 day) that the H2O2 abundance can be considered to be photochemically controlled. However, H2O2 should still be expected to show significant temporal and spatial variability through its sensitivity to transport-related fluctuations in O3, NOy(=NO+NO2+HNO3+HNO4), and H2O. From the available observational data on short-term variability in these longer-lived species, H2O2 concentrations should vary temporally four- or fivefold at altitudes between 15 and 35 km. This concentration range about the model profile includes two recently reported upper limits for mid-stratospheric H2O2. The earlier tentative identification of stratospheric H2O2 by Waters et al. (1981) is somewhat larger than the expected upper limit of variation based on a one-dimensional transport-kinetics model. Several implications for stratospheric observation of H2O2 arise from its expected variability. Model-predicted H2O2 abundances are very sensitive to important kinetic parameters, but small sets of H2O2 measurements alone will not be useful to test model behavior. In concert with simultaneous measurements of O3, H2O, and NOy, H2O2 abundances could be use to closely verify model treatment of HOx(=HO+HO2) photochemistry. Finally, the HOx abundance and its partitioning between HO and HO2 are quantities of great importance in the chemistry of the mid and lower stratosphere. Inference of these values from H2O2 and some other independent quantity related to HOx involves knowledge of only a small set of photochemical parameters and could supplement direct observation of the radical species.