Abstract Tectonic interpretation of the central Sierra Nevada—whether the crest of the Sierra Nevada (California, USA) was uplifted in the late Cenozoic or whether the range has undergone continuous down-wearing since the Late Cretaceous—is controversial, since there is no obvious tectonic explanation for renewed uplift. The strongest direct evidence for late Cenozoic uplift of the central Sierra Nevada comes from study of the Trachyandesite of Kennedy Table, which followed the course of the Miocene San Joaquin River but has a steeper gradient than the modern river. Early workers attributed this steeper gradient to tilting of the Sierra Nevada block since the late Miocene, resulting in 2 km of range-crest uplift. However, this interpretation has been contested on grounds that the Miocene river gradient had to be assumed and that the Sierran Batholith could have warped during tilting, thus failing to uplift the range crest. The objective of this study was to obtain quantitative data that test these criticisms. The Trachyandesite of Kennedy Table is a chain of 33 remnants of a single lava flow as thick as 65 m, preserved for 21 km from Squaw Leap to Little Dry Creek, close to the modern San Joaquin River in the foothills of the Sierra Nevada. Several remnants lie on fluvial gravel of the late Miocene San Joaquin River. Early workers speculated that the lava concealed its own (unrecognized) vent, but in 2011, we identified the vent on the Middle Fork of the San Joaquin River, 13.5 km south of Deadman Pass and 70 km northeast of Kennedy Table. The vent complex intrudes Cretaceous granite, has 285 m relief, and is an intricately jointed intrusion that grades up into a glassy lava flow. Composition (58% SiO2) and 40Ar/39Ar age (9.3 Ma) are identical at the vent and downstream. Basal elevations of remnants were recorded, and the present-day basal gradients of several were adjusted for apparent dip and projected along a vertical plane at 220° (the estimated tilt azimuth). The basal gradients are far steeper than that of the modern river, but they differ slightly from reach to reach and are thus inconsistent measures of the post-Miocene tilt. Likewise, relief eroded atop most remnants renders modeling of upper surfaces suspect. At Little Dry Creek, however, a chain of nine remnants rests on fluvial floodplain sand and gravel; this chain trends 230°, and its smooth basal contact now dips 1.36° (adjusted at 220°). Projection of this dip 89 km from the 207 m base of the most distal remnant at Little Dry Creek to the vent intrusion falls far below the 2760 m intrusion-to-lava-flow transition near the Sierran crest, showing that the Sierran block has not undergone pronounced convex warping. Using elevation data on paleoriver meanders preserved by the lava flow, we show that the paleogradient has a cosine dependence on meander-section azimuth, indicating tilting. Subtraction of 1.07° of dip restores the data to an azimuth-independent configuration, indicating total tilting since 9.3 Ma of 1.07° and an original large-scale gradient of 0.46°, similar to the published value of 0.33° at Squaw Leap, but larger than the previously obtained value of 0.057° at Little Dry Creek. Subtraction of those Miocene estimates from the observable 1.643° tilt along the section from Little Dry Creek to the vent yields vent uplift of 2464 m (for 0.057°), 1835 m (for 0.46°), and 2040 m (for 0.33°). Confirmation of earlier assumptions regarding Miocene river gradient and block rigidity greatly strengthens the case for ~2 km of late Cenozoic uplift of the central Sierra Nevada crest.