The increasing coverage of continuous GPS stations along the Cascadia Margin, together with improvements in GPS data analyses are providing not only an unprecedented research tool for the study of regional neotectonics, but also an effective methodology for improving estimates of seismic hazard. Precise GPS observations allow direct detection of elastic strain buildup which controls both spatial and temporal occurrences of large earthquakes, exactly those seismic events which pose the largest hazard but whose recurrence intervals and spatial distributions are not well defined by our short historical records. The surface coseismic displacements recorded at continuous GPS stations for the Nisqually (M=6.8) earthquake clearly demonstrate that GPS can also be used to study the hazard associated with large ruptures deep in the subducting slab. Coseismic GPS observations can help to define the fault plane and the redistribution of the Coulomb stress, which may control the occurrence of subsequent earthquakes. The recent discovery of aseismic slip on the deeper Cascadia subduction interface suggests that stress accumulation across the locked megathrust zone can occur in discrete pulses. Such slip events could evolve into a trigger mechanism for a great thrust earthquake. Consequently, the identification and monitoring of deep aseismic slip through continuous GPS observations may lead to a time-variant estimate of seismic potential of the subduction megathrust.