Gravity and magnetic maps of western Washington reveal the lateral structure and fabric of the Washington Coast Range, Puget Basin, and southern Washington Cascade Range. The magnetic and gravity maps show large amplitude positive anomalies associated with the shallow but largely buried section of Washington Coast Range mafic rocks which are separated by negative anomalies over deep sedimentary basins. The positive anomalies indicate that the Coast Range mafic basement extends farther east than previously thought, at least as far east as the longitude of Seattle. Linear and steep gravity and magnetic gradients indicate many unmapped, often buried faults in the Washington Coast Range Province. Magnetic highs are also associated with mapped batholiths in the Cascade arc. Several magnetic highs observed east of the Coast Range rocks and west of these batholiths may be associated with buried Tertiary plutons or ophiolites. Two-dimensional gravity and magnetic modeling constrained with geological and other geophysical data indicate that the Coast Range Province rocks are about 1 km thick at the coast, thickening to as much as 30 km near their postulated eastern edge. A maximum boundary on the average density of the upper 15--20 km of the rocks that compose the Coast Range Province of 2920 kg/m3 was established by the modeling, suggesting a composition largely of basalt and gabbro with little interbedded sediments. Under these rocks may be mantle or a subduction complex composed of dense mafic, ultramafic, and sedimentary rocks like that proposed to underlie Vancouver Island. Previous gravity models of the Washington margin include lower densities for the proposed subduction complex than for Vancouver Island, suggesting a lower component of mafic and ultramafic rocks than the rocks underlying Vancouver Island. However, my Washington model requires that the proposed subduction complex be more dense than the trench sediments and, therefore, that material denser than sediments be incorporated within it. The absence of continental mantle and the modeled wedge shape of the Coast Range Province upper crust suggest that erosion of the bottom of the overriding plate by subduction processes may have occurred.