In-field sampling programs generally take time, equating to money, and unless you have an onsite laboratory, you have the issue of having to wait long periods for your data. Until recently, field portable X-ray fl uorescence (FPXRF) analysers, which can help provide instant indicative results, were not able to detect penalty elements, such as Al and Si, or do so reliably enough to act upon.The purpose then of this field test then was to show that the latest FPXRF devices are capable of producing reliable readings on field samples of variable grain size, in particular for Al and Si, such that the readings still correlate well with the assay for those samples and thus be reliable for an iron ore grade control application.We tested RC chip samples on-site in the Pilbara, Western Australia, with the Niton, analysing for a range of elements including Fe, Mn, Al and Si and then compared the results for those elements against assay data for the same. A strong correlation would tell us that the FPXRF results were reliable. Furthermore, to show the effect of simple sample preparation, we also analysed pulps of the same samples in the same way.The correlations were good for the all of the four elements of interest for the RC chip samples (R2 values between 0.82 and 0.9), and excellent for the pulps (R2 values between 0.89 and 0.998).The results also showed a direct correlation between sample preparation and improved quality of results with the FPXRF.The data showed allowed for two conclusions:1. the FPXRF can produce data from unprepared RC chips for Al and Si that, when compared with assay, is reliable for a fi eld application in drilling and grade control; and2. the greatest limitation to data quality is sample preparation.