Variability in ore feed grade and mineralogy can affect flotation stability, reagent demand and metallurgical performance. At Kanmantoo, crusher feed is prepared by reclaiming material from runof-mine stockpiles according to a planned blend based on geological information. While this approach provides a target feed grade, the actual sequence of material delivered to the crusher can be affected by ore heterogeneity, stockpile segregation and reclaim practice. Until recently, there was no continuous elemental measurement between the crusher and the coarse ore stockpile to show how closely the realised feed matched the planned blend. A Thermo Scientific cross-belt analyser using prompt gamma neutron activation analysis (PGNAA) was installed on the crushing circuit discharge conveyor feeding the coarse ore stockpile (COS). The analyser provides continuous bulk measurement of copper, sulfur and other elements before the material enters the final stockpile buffer. A post-commissioning study was undertaken to determine whether this measurement could provide useful information for ore-blending control. Factory and site validation were used to confirm analytical performance. Minute-level measurements from the cross-belt analyser and the AnStat-430, a dedicated flotation-feed X-ray florescence (XRF) slurry analyser, were then used to establish the effective process response between crusher product and flotation feed. This timing relationship was subsequently applied when comparing the planned blend estimate, the realised cross-belt analyser measurement, the slurry analyser measurement and independent reconciled feed shift composite assays. The commissioning results show that the planned blend average does not fully describe either the realised crusher-product grade or the variability within a blend. Grade changes measured by the cross-belt analyser remained evident downstream after stockpile and milling attenuation, while independent shift composite assays provided further confirmation that the upstream measurement was representative of the material subsequently processed. The study demonstrates that the cross-belt analyser provides a timely and process-relevant measurement suitable for supporting ore-blending decisions at Kanmantoo. The next stage is to use the measurement in a controlled operator-assisted blending trial to determine whether active response to the measured grade and variability can deliver repeatable improvements in feed stability.