The flow of plasma along open field lines at high latitudes is highly variable and depends both on conditions in the underlying ionosphere and thermosphere and on the transport of particles and energy from the magnetosphere. Past attempts to model this time variability have, for the most part, examined the response of the plasma on a stationary field line to certain prespecified boundary conditions and heat sources. While such prespecified conditions may bear some resemblance to what occurs naturally, they are artificial and cannot be expected to yield a truly quantitative understanding of the various physical processes that interact to produce the dynamic polar wind. The present study is one in a series of studies that attempts to eliminate this artificiality by coupling the mathematical description of the polar wind to a three-dimensional time-dependent model of the high-latitude ionosphere. In this study, an individual flux tube of plasma is followed as it moves under the influence of combined corotation and convection electric fields. Boundary conditions at the lower end of the flux tube are obtained from the ionosphere model, which takes into account all significant particle species, chemical reactions, and heat sources that contribute to the state of the ionosphere. A multi-ion macroscopic particle-in-cell code is used to model the plasma in the flux tube. A description of the behavior of H+ and O+ for the altitude range from 2000 km to about 8 Earth radii is obtained as the flux tube moves along the trajectory, which traverses regions of the subauroral ionosphere, dayside and nightside ovals, and polar cap. The goal of the study is to determine the extent to which ion trapping can occur in the polar wind and the effects that collisions, wave-particle interactions, centrifugal acceleration, and varying ionospheric conditions have on the trapped ions. The main conclusion of the study is that O+ trapping is important and it acts to increase the O+ density at high altitudes. ¿ 1998 American Geophysical Union