For groundwater studies, the collection of sediment geochemistry is often beyond the scope, and budget of many programs, and is generally not included as part of routine data collection. Portable x-ray fluorescence spectrometer (pXRF) has proven to be a successful tool to characterize the chemostratigraphy of glacial derived materials collected from boreholes in southern Ontario. Portable XRF provides near-total geochemistry results similar to fusion and multi-acid methods for 14 elements with minimal sample preparation and at low cost. An extensive suite of near surface samples provides characterization of the regional geochemistry. In a collaborative project with the OGS and the GSC two transects are being completed (E-W, N-S) to provide a framework of subsurface geochemistry. To eliminate the effects of variability in sample grain-size, sample volume, and to minimize nugget effects, samples are dried and sieved to <0.063 mm (silt + clay) prior to analysis. To ensure quality control, a number of Standard Reference Materials (SRM) and Certified Reference Materials (CRM) are analyzed with each project, and comparisons made with previously published results. For further quality control, a sub-suite of sediment samples are analyzed by ICP-MS/ES using lithium borate fusion, multiacid, and aqua regia digestions. Bivariate plots comparing pXRF to ICP-MS/ES display a high degree of linearity (r2 > 0.8) for Ca, Fe, K, Mn, Rb, Sr, V, Zn, Zr, and to a lesser degree for Ba, Cu, Cr, Ti, and Pb. These 14 elements return precise and generally accurate results within each borehole; however, continued analyses of CRM's and SRM's has demonstrated display drift in accuracy between projects. Resulting data for meet the US EPA criteria for quantitative data quality based on r2 values and y=mx+b relationships. Concentration levels play a significant factor in the accuracy of the pXRF data. At low concentration levels near the detection limit of the pXRF, there can be greater scatter in results. At high concentration levels, data needs to be adjusted using post-data calibration to obtain accurate results.