We employ a three-dimensional ring current model to trace the bounce-averaged drift of singly charged ions during storm-associated enhancements in the convection electric field. Using the simulation results, we map proton phase space density during the main and recovery phases of a storm in accordance with conservation of phase space density f. We map from an initial quiescent phase space distribution that is obtained by solving the steady state transport equation (bounce-averaged charge exchange balancing bounce-averaged radial diffusion) with observed plasma sheet proton spectra as outer boundary conditions. We obtain proton pitch angle distributions at L~3--4.5 by evaluating f at representative ring current energies (~20--170 keV). We find that the prestorm and stormtime proton pitch angle anisotropy at any given L between 3 and 4.5 increases with particle energy in agreement with observations. The actual anisotropy at specific energies depends strongly on the shape of the plasma sheet source spectrum at ~0.5--3 keV. Relatively large enhancements in the stormtime phase space density from the quiescent distribution occurs at all pitch angles for low energies (≲80 keV) except where the quiescent distribution lies on open drift shells. These increases result primarily from stormtime access to L≲4.5 along open drift trajectories from the plasma sheet. For higher-energy (≳150 keV) protons, which are transported via radial diffusion, there is little change in the anisotropy over a 3-hour storm. At intermediate energies (E~80--150 keV) the stormtime enhancements in the phase space density can vary quite strongly with equatorial pitch angle. For such particles the stormtime transport is intermediate between directly convective and quasi-diffusive, which can complicate the analysis of stormtime pitch angle anisotropies at intermediate energies. Decay lifetimes obtained from CRRES data during the recovery phase of a moderate storm are found to be considerably shorter than charge-exchange lifetimes. It thus appears that charge exchange alone is not enough to explain the observed rapid decay of the proton ring current. ¿ 1998 American Geophysical Union