Daily average solar proton flux data for the years 1963–1984 (two solar cycles) have been used in a proton energy degradation scheme to derive ion pair production rates and, subsequently, HOx (H, OH, HO2) and NOx (N, NO, NO2) production rates. These HOx and NOx production rates are computed in a form suitable for inclusion in an atmospheric two-dimensional time-dependent photochemical model. The HOx increases, although large for certain solar proton events (SPEs), are relatively short-lived because the HOx species have lifetimes of only hours in the middle atmosphere. For longer-lived NOx species, increases are important for 2–4 months past the more intense SPEs but are generally negligible 6 months after the SPE. The only exception to this scenario was the gigantic August 1972 SPE, whose stratospheric effects lasted about a year past the event. Comparisons of model results with the ozone data from the Nimbus 4 backscattered ultraviolet (BUV) instrument indicate relatively good agreement in the time dependence and magnitude of the ozone depletion for the middle stratosphere between the model and measurements for the August 1972 SPE and for 2 months past the event. The model predictions for the August 1972 SPE indicate at most a 1% decrease in total ozone at the highest latitudes with a significant interhemispheric difference. The model predicts a larger middle latitude stratospheric ozone change in the southern than the northern hemisphere caused by the difference in seasons between the two hemispheres. The computed ozone decreases associated with the HOx and NOx increases are substantial in the upper stratosphere at high latitudes for only a few SPEs in the 22 years studied. A mechanism is presented for transport of NOy from the stratosphere to the ground, which may be involved in the enhancements in nitrate fluxes noticed in Antarctic deposition data. Our computations, however, indicate that the SPE contributions to these nitrate fluxes (even for the August 1972 SPE) are probably small.