The shape of the dayside magnetopause has been studied from both a theoretical and an empirical perspective for several decades. Early theoretical studies of the magnetopause shape assumed an inviscid interaction and normal pressure balance along the entire boundary, with the interior magnetic field and magnetopause currents being solved self-consistently and iteratively, using the Biot-Savart Law. The derived shapes are complicated, due to asymmetries caused by the nature of the dipole field and the direction of flow of the solar wind. These models contain a weak field region or cusp through which the solar wind has direct access to the ionosphere. More recent MHD model results have indicated that the closed magnetic field lines of the dayside magnetosphere can be dragged tailward of the terminator plane, so that there is no direct access of the magnetosheath to the ionosphere. Most empirical studies have assumed that the magnetopause can be approximated by a simple conic section with a specified number of coefficients, which are determined by least squares fits to spacecraft crossing positions. Thus most empirical models resemble more the MHD models than the more complex shape of the Biot-Savart models. In this work, we examine empirically the effect of the cusp regions on the dayside magnetopause, and we test the accuracy of these models. We find that during periods of northward IMF, crossings of the magnetopause that are close to one of the cusp regions are observed at distances closer to Earth than crossings in the equatorial plane. This result is consistent with the results of the inviscid Biot-Savart models and suggests that the magnetopause is less viscous than is assumed in many MHD models. ¿ American Geophysical Union 1995