Bathymetric data reveal abundant submarine landslides along the deformation front of the northern Cascadia margin that might have significant tsunami potential. Radiocarbon age dating showed that slope failures are early to mid-Holocene. The aim of this study is the analysis of slope stability to investigate possible trigger mechanisms using the factor of safety analysis technique on two prominent frontal ridges. First-order values for the earthquake shaking required to generate instability are derived. These are compared to estimated ground accelerations for large ( M =5 to 8) crustal earthquakes to giant ( M =8 to 9) megathrust events. The results suggest that estimated earthquake accelerations are insufficient to destabilize the slopes, unless the normal sediment frictional resistance is significantly reduced by, for example, excess pore pressure. Elevated pore pressure (overpressure ratio of 0.4) should significantly lower the threshold for earthquake shaking, so that a medium-sized M =5 earthquake at 10 km distance may trigger submarine landslides. Preconditioning of the slopes must be limited primarily to the mid- to early Holocene as slope failures are constrained to this period. The most likely causes for excess pore pressures include rapid sedimentation at the time of glacial retreat, sediment tectonic deformation, and gas hydrate dissociation as result of ocean warming and sea level rise. No slope failures comparable in size and volume have occurred since that time. Megathrust earthquakes have occurred frequently since the most recent failures in the mid-Holocene, which emphasizes the importance of preconditioning for submarine slope stability.