The rare earth elements (REEs) are critical metals for developing renewable technologies. Regolith- hosted deposits are emerging sources of REEs, though the economic viability of these ores depends on their hosting significant volumes of mineralised material which compensate for their low metal grades (<2000 ppm eree). the clay-dominated nature of regolith-hosted ree deposits, which contributes to the difficulty of distinguishing transported cover units from in situ weathered units, makes their delineation through traditional methods like petrological analysis and core logging challenging. to expedite deposit-scale characterisation of the regolith-hosted splinter rock ree prospect, western australia, we applied principal component analysis and k-means clustering to>3000 whole-rock geochemical assays of its mineralised and barren regolith horizons. K-means analysis successfully separated all samples into five stratigraphically continuous horizons (three barren and two mineralised). The horizons were interpreted as units of sedimentary cover and weathered granites, respectively. Once clustered, the mineralogical and metallurgical features of all samples were extrapolated from select drill cores defined by hyperspectral analysis. This facilitated the identification of several robust mineralogical, geochemical and metallurgical trends at the deposit scale: . The highest REE grades exist in the granitic saprolite and saprock. . Relative to the light REEs, the heavy REEs are enriched at the saprolite-saprock boundary and become depleted with increasing depth in the saprock. . Optimal metallurgical conditions occur near the saprolite-saprock interface. . accumulation of the economically- and environmentally-important 'magnet' REEs occurs mainly in the saprock. . MagREE enrichment can be linked to the development of negative Ce anomalies. Principal component analysis also facilitated the establishment of tailored geochemical ratios for classifying future samples into their appropriate clusters/horizons, such that this study highlights the multivariate statistical analyses of geochemical data as an efficient method for outlining deposit- scale trends and zones of economic metal enrichment in regolith-hosted REE deposits.