Models of magnetic and gravity anomalies along two E-W transects offshore central Oregon, one of which is coincident with a detailed velocity model, provide quantitative limits on the structure of the subducting oceanic crust and the crystalline backstop. The models indicate that the backstop-forming western edge of the Siletz terrane, an oceanic plateau that was accreted to North America ~50 million years ago, has a seaward dip of less than 60 ¿3. Seismic, magnetic, and gravity data are compatible with no more than 2 km of subducted sediments between the Siletz terrane and the underlying crystalline crust of the Juan de Fuca plate. The data also suggest the presence of a N-S trending, 200-km-long basaltic ridge buried beneath the accretionary complex from about 43 ¿N to 45 ¿N. Although the height and width of this ridge probably vary along strike, it may be up to 4 km high and several kilometers wide in places and appears to be locally in contact with the Siletz terrane beneath Heceta Bank. Several models for the origin of this ridge are discussed. These include: a silver of Siletz terrane detached from the main Siletz terrane during a late Eocene episode of strike-slip faulting; imbrication and thickening of subducted oceanic crust in place; an aseismic ridge rafted in on the subducting oceanic crust during the past 1.2 million years; and a series of ridges and/or seamounts rafted in over a longer period of time and transferred from the subducting plate to the overlying plate. The last model is the most consistent with the complicated history of local uplift, subsidence, and slope instability recorded in the ridges, basins, and banks of this part of the margin. We speculate that the massive seaward dipping western edge of the Siletz terrane in this region inhibits subduction of seamounts and sediments, resulting in formation of buried ridge as the accumulated flotsam and jetsom of subduction. This process may also be responsible for thickening of lower accretionary complex material, oversteepening of slopes leading to massive slumping, and north-south extension through strike-slip faulting in the accretionary complex to the west of the buried ridge. Regardless of its origin, the ridge may currently be acting as an asperity inhibiting subduction. ¿ 1999 American Geophysical Union