Ombrotrophic peatlands are recognized archives of past atmospheric mineral dust deposition. Net dust deposition rates, grain size, mineral hosts and source areas are typically inferred from down-core elemental data. Although elemental analysis can be time efficient and data rich, there are some inherent limitations. X-ray diffraction (XRD) analysis allowsdirect identification of mineral phases in environmental samples but few studies have applied this method to peat samples and a well-developed protocol for extracting the inorganic fraction of highly organic samples (>95%) is lacking. We tested and compared different levels of pre-treatment: no pre-treatment, thermal combustion (300, 350, 400, 450, 500 and 550 degrees C) and chemical oxidation (H2O2 and Na2S2O8) using a homogenised highly organic (>98%) composite peat sample. Subsequently, minerals were identified by XRD. The results show that combustion is preferred to chemical oxidation because it most efficiently removes organic matter (OM), an important pre-requisite for identifying mineral phases by XRD analysis. Thermally induced phase transitions can be anticipated when temperature is the only factor to take into consideration. Based on the data required in this studythe recommended combustion temperature is 500 degrees C which efficiently removes OM while preserving a majority of common dust minerals.