Twenty-five Holocene sand samples were collected from the mouths of drainage basins or from washes that led to homogeneous or mixed sources in the mountains adjacent to Salton Basin for detailed petrologic study. Bulk sand-sized and medium-grained fractions separated from each sample were consolidated with epoxy, thin-sectioned, and stained for K-feldspar and plagioclase. Thin-sections made from bulk fractions were point counted following the Gazzi-Dickinson method. Thin-sections made from medium-grained fractions were point counted for Qmnu, Qmu, Qp (2-3) and Qp (>3). In addition, during the bulk fraction count, a running tabulation of aphanitic rock fragments was kept. The results of this study suggest that the average Holocene sand derived from mountains marginal to Salton Basin contain 44.6% ± 11.3% Q, 49.6% ± 9.5% F, and 5.7% ± 7.9% L. On the QFL provenance-discrimination diagram developed by W. R. Dickinson and associates, 72% of the samples plot in the basement-uplift field and, therefore, accurately reflect their derivation from the coarse-grained metamorphic/plutonic roots of a segment of the Cordilleran magmatic arc. As might be expected, two samples derived from unmetamorphosed volcanic sources plot about the boundary between the transitional-arc and dissected-arc subfields. The remaining five samples, however, extend into the transitional-continental and recycled subfields, probably as a result of mixing plutonic and/or metamorphic detritus with sand derived from quartz-rich metamorphic rocks or from older Neogene sedimentary material. Medium-grained quartz populations are complex, and plot in the middle-to upper-rank and low-rank metamorphic fields on the QmnuQmuQp discrimination diagram developed by A. Basu and colleagues. This result accurately reflects the deep crustal origin of the bedrock in the mountains adjacent to Salton Basin, and its subsequent uplift and deformation in large scale dextral shear zones associated with the San Andreas and related fault systems. Data obtained in this study generally yield internally consistent results which accurately reflect the known provenances and tectonic settings of the samples studied. Therefore, I recommended that petrologists utilize the QmnuQmuQp and QFL provenance-discrimination models when attempting to unravel the provenance of ancient feldspathic sandstones. However, I also caution them that local source areas can produce sands whose compositions, when plotted on the QmnuQmuQp or QFL provenance-discrimination diagrams, may lead to an erroneous or overly simplistic interpretation of source rock characteristics or tectonic setting. In addition, data described here also indicate that the QFL, medium-grained quartz, and aphanitic-rock-fragment populations of sand derived from the mountains adjacent to Salton Basin are unlike those of Holocene Colorado River sand. Distinct compositional differences between the two groups of sand should enable future workers to better recognize and map ancestral Colorado River sand in the Salton Basin. Recognition of displaced Colorado River sand would place constraints on the movement history along the San Andreas and associated strike-slip fault system and, hence, on the origin of the Salton Basin-Gulf of California rift system.