Least cost paths (LCPs) have been widely used by archaeologists and geographers to reconstruct probable routes of movement within past landscapes using GIS-based modelling. By definition, LCP modelling is based on the premise that 'least cost' features as the primary factor in the decision-making of movement. It can be argued, however, that movement within the landscape was structured not only by the need to minimize costs but also by the desire to maximize benefits. This study introduces a new way of estimating terrain costs that can factor in this tendency towards 'benefit maximization' when modelling routes. This alternative methodology features a distinctive way of perceiving the landscape, which differs from the grid-based division of the landscape generally used in LCP modelling. The landscape is seen to be made up of ten different 'hillslope position units' (which are generic landscape units widely used in landscape evolution studies) and 'movement suitability' values are estimated for each of the hillslope position units. These values are then used to produce a 'hillslope position unit movement suitability' (HPMS) cost surface. The evaluation of the HPMS cost surface is undertaken by comparing HPMS-based routes and slope gradient-based LCPs against the historical roads of Korea's Joseon Dynasty . It is observed that HPMS-based routes demonstrate a greater degree of correspondence with historically confirmed routes, compared to slope gradient-based LCPs. The similarity between HPMS-based routes and historically confirmed routes is found to be greater in mountainous regions vis-à-vis non-mountainous regions. It is proposed that, by taking into account human desire to utilize beneficial landscapes, HPMS-based cost surfaces may result in modelled routes that are closer to actual past experiences.