Forest management practices often result in significant changes to hydrologic and geomorphic responses at or near the earth's surface. A well-known, but not fully tested, hypothesis in hillslope hydrology[sol ]geomorphology is that a near-surface permeability contrast, caused by the surface compaction associated with forest roads, can result in diverted subsurface flow paths that produce increased up-slope pore pressures and slope failure. The forest road focused on in this study is located in a steep forested, zero-order catchment within the H. J. Andrews Experimental Forest (Oregon). A three-phase modelling effort was employed to test the aforementioned hypothesis: (i) two-dimensional (vertical slice), steady-state, heterogeneous, saturated subsurface flow simulations at the watershed scale for establishing the boundary conditions for the catchment-scale boundary-value problem in (ii); (ii) two-dimensional (vertical slice), transient, heterogeneous, variably saturated subsurface flow simulations at the catchment scale for estimating near-surface hydrologic response and pore pressure distributions; and (iii) slope stability analyses, using the infinite slope approach, driven by the pore pressure distributions simulated in (ii), for assessing the impact of the forest road. Both observed and hypothetical rainfall events are used to drive the catchment-scale simulations. The results reported here support the hypothesis that a forest road can have an effect on slope stability. The permeability contrast associated with the forest road in this study led to a simulated altering of slope-parallel subsurface flow with increased pore pressures up-slope of the road and, for a large rainfall event, a slope failure prediction. Copyright © 2005 John Wiley & Sons, Ltd.