It is well known that convection electric fields have an important effecton the ionosphere at high latitudes and that a quantitative understanding of their effect requires a knowledge of plasma convection over the entire high-latitude region. Two empirical models of plasma convection that have been proposed for use in studying the ionosphere are the Volland and Heelis models. Both of these models provide a similar description of two-celled ionospheric convection, but they differ in several ways, in particular, in the manner in which plasma flows over the central polar cap and near the polar cap boundary. In order to obtain a better understanding of the way in which these two models affect the ionosphere, two separate runs of our high-latitude, time-dependent ionospheric model were made, with only the convection models distinguishing the two runs. It was found that the two models lead to differences in the ionosphere but often the differences are subtle and are swamped by universal time effects. The most notable differences are in predictions of the height of the F2 peak and in the ion temperature, particularly along the evening polar cap boundary and in the cusp region. For these two parameters, the differences caused by the two different convection models dominate the universal time effects. One question that arises is whether one could examine measurements of plasma density and temperature and determine which of the two convection models most accurately represents actual ionospheric convection. Unfortunately, it is expected that when the effects of other ionospheric inputs are considered, such as the neutral wind, the uncertainties are sufficiently large that the characteristic differences between the Volland and Heelis convection models cannot be clearly identified in an examination of plasma density and temperature measurements.