A distributed shallow groundwater model related to slope stability is described to assess the spatial distribution of pore water pressure in steep forested terrain in British Columbia. Additionally, effects of timber harvesting and roads on measured changes in pressure head during rainstorms were evaluated for the first time to assess the need for incorporating different hydrological components in the event-driven distributed model. Although explicit spatial quantification of pore water pressure requires many measurements for accurate prediction, model performance using average parameter values was reasonable when compared with pressure heads measured at nine spatially distributed sites. Increases in maximum pressure head (varying from 9 to 28 cm) between preharvesting (after road construction) and postharvesting rainstorm events were observed in seven of nine sites. The remaining two sites showed either a small decrease (≈5 cm) or similar peak pressure heads following harvesting. Peak pressure head evaluated at one piezometer located 46 m downslope of the road decreased substantially (≈50 cm) after road construction during moderate rainstorms and then recovered following harvesting. Piezometric responses in sites upslope of the road were not affected by road construction but did increase after harvesting. Moderate storms caused the largest relative increases in pressure head between preharvesting (after roads) and postharvesting conditions; such increases were small during large storms, lending support to the idea that timber harvesting in temperate forests enhances hydrologic response only during small and moderate storms. Since landslides in coastal Pacific Northwest are typically caused by large winter rainstorms, it appears more justified to include better spatial representation of soil physical and engineering parameters in the distributed shallow groundwater model compared to specifying evapotranspiration; road hydrology may, however, need to be included.