Although it has become well established that the low-altitude polar cusp moves equatorward during intervals of southward interplanetary magnetic field (IMF Bz<0), many other important aspects of the cusp's response to IMF components are not as well investigated. An algorithm for identifying the cusp proper was applied to 12,569 high-latitude dayside passes of the DMSP F7 satellite (which is in a nearly circular polar orbit at ~838 km altitude), and the resulting cusp positioning data were correlated with the IMF (IMF data were available for about 25% of the cases). It was found that the peak probability of observing the cusp shifts prenoon for By negative (positive) in the northern (southern) hemisphere and postnoon for By positive (negative) in the northern (southern) hemisphere. The By induced shift is much more pronounced for southward than for northward Bz, a result that appears to be consistent with elementary considerations from, for example, the antiparallel merging model. No interhemispherical latitudinal differences in cusp positions were found that could be attributed to the IMF Bx component. As expected, the cusp latitudinal position correlated reasonably well (0.70) with Bz when the IMF had a southward component; the previously much less investigated correlation for Bz northward proved to be only 0.18, suggestive of a half-wave rectifier effect. The ratio of cusp ion number flux precipitation for Bz southward to that for Bz northward was 1.75¿0.12. The statistical local time (full) width of the cusp proper was found to be 2.1 hours for Bz northward and 2.8 hours for Bz southward. ¿ American Geophysical Union 1989