Simple, approximately self-consistent models of magnetic recording as well as materials limitations are used to show that, ultimately, vertical recording will have no fundamental advantage in signal strength, bit shift, or resolution over conventional longitudinal recording. While very narrow transitions are in principle stable when

is vertical, this analysis predicts written transitions that are much broader than the demagnetizing limit, just as in longitudinal recording. The limiting factor is shown to be the gradient of the writing head\´s field.