In the Darling Range of Western Australia, lateritic bauxite profiles are usually formed from granite and meta-dolerite. Complex assemblages of regolith materials are the result of generally residual in situ isovolumetric chemical and physical weathering of these diverse rocks. Bauxite mining by Alcoa World Alumina Australia removes the upper 4 to 6 m of the resultant profile. Regolith samples collected from pre-mining drill-holes are analysed using Fourier Transform Infrared (FTIR), microwave digest and X-ray fluorescence procedures to determine the grade and depth of bauxite ore (available alumina greater than 27.5%). The percentages of available alumina (Al), reactive silica (Si), total silica (ST), total iron (Fe) and the ratio of (Si/ST) for samples collected from below the depth of mining have been used to predict regolith type. FTIR regolith maps of bauxite mine floors can be produced. Known regolith material was analysed that is FTIR based using multivariate cluster analysis (Wards Method) and regolith codes allocated to cluster groups with a purity of 75 to 90% for preserved doleritic fabric, clay rich and iron oxide cemented regolith. The mean and standard deviation within each cluster group was then used to construct a classification tree of diagnostic variables. Appling the classification tree to data from drill-holes at two bauxite mine pits enabled the identification of zones of clay-rich, iron oxide-cemented and doleritic fabric materials. Remaining zones generally contained quartz-rich regolith which requires deep (1.5 m) ripping to fracture the dense, hard regolith that is limiting to plant root growth. Rehabilitation management can now target regolith type and FTIR regolith maps produced allow allowing early development of mine pit closure plans before mining commences. Ripping, tree and understory seeding strategies can be adjusted to regolith type.