The Sierra Nevada of California has experienced several kilometers of uplift in the last 10 m.y., but its present elevation is isostatically compensated by a crustal root that has probably existed since Late Cretaceous. Flexural isostasy can reconcile these observations: Basin and Range faulting and extension starting 10 m.y. ago allowed rapid uplift of an eroded and regionally compensated magmatic arc underlain by an excess crustal root. A quantitative model of an unbroken elastic plate buoyed up by the excess root provides an explanation for the estimated 10-m.y. paleotopography of the range. The same model fits the present topography when broken (shear and fiber stresses not transmitted) at the location of the Owens Valley fault zone. The broken-plate model predicts the appropriate amount of westward tilt of the Sierra Nevada and can explain the pronounced eastward tilt of the White and Inyo mountains. The position where paleo and present drainages in the Sierra cross over in elevation is the most sensitive indicator of the elastic thickness required of the lithosphere. An effective elastic thickness of 50 km best fits the observed cross-over distances and is consistent with the amount of post-10 m.y. uplift. Calculated fiber and shear stresses suggest that actual extension of the crust is more important in permitting the uplift than the mere presence of normal faults. Part of the residual isostatic gravity anomaly pattern (high gravity over the eastern Great Valley and western Sierra Nevada, low residual gravity over the high Sierra) may be explained by the continuing flexural support of departure from local isostatic equilibrium. The model implies that before 10 m.y. ago the Sierra should have been marked by a negative isostatic gravity anomaly of more than -100 mGal, flanked by isostatic gravity highs. Such large isostatic anomalies have not been observed to data over eroded arcs, but the Sierra has uniquely low heat flow and, presumably therefore, high lithospheric strength, and the Sierran root is at the upper limit of size that could be isostatically suppressed. ¿ American Geophysical Union 1988