K2O contents have long been recognized as a potential indicator of tectonic processes, and in the Sierra Nevada, California, high-K2O volcanism has been attributed to lithosphere root delamination. However, new data from the central Sierra suggest a very different control: K2O concentrations can be explained by variations in the degree of partial melting in the mantle, where high-K2O volcanics are derived from low-degree partial melts of mantle lithosphere. Field evidence in the central Sierra further suggests that the pulse of high-K2O volcanism there was synchronous with the development of a pull-apart structure along a series of right-stepping dextral transtensional faults at the onset of Walker Lane transtensional faulting. In our alternative interpretation, high-K, low-degree partial melts were tapped by the inception of transtensional stresses, recording the birth of a plate boundary. We speculate that high-K2O lavas in the southern Sierra are similarly related to the onset of transtensional stresses, not delamination. A regional southward increase in incompatible element contents and decrease in erupted volumes are also consistent with a model for crustal thickness controls on magmatism. Depth-integrated density models show that dry mafic magmas beneath thick crust have insufficient buoyancy to erupt, but low-degree partial melts carry sufficient volatiles to allow eruption; as with K2O, degree of partial melting, not source-region heterogeneity, controls water contents and buoyancy.