A novel approach was developed to examine coke abrasion resistance in-situ at temperatures of up to 950 °C and in a controlled gas atmosphere using rotational tribological testing. The originality of this approach lay in the ability to apply tribological testing to the porous coke surface at elevated temperatures under a controlled inert or CO2 reactive atmosphere. Coke wear characteristics were quantified via (i) the application of advanced microscopy and image analysis techniques, and (ii) analysis of the coefficient of friction (COF) during tribological testing. Two pilot oven cokes were examined in this study, which were generated from single coking coals of similar rank but different petrographic composition. The cokes tested showed a deterioration in abrasion resistance even at 400 °C, and this was accentuated at 950 °C. The COF and the severity of the wear as a function of temperature and gas atmosphere were markedly different between the two cokes. Coke C396 from a mid-to-low vol, medium vitrinite coal showed no statistically significant differences between the wear severity results recorded following room temperature testing and high temperature testing in a CO2 environment. Conversely, the more reactive coke C426 with a higher parent coal inertinite content showed clear differences in the extent of abrasive wear at room temperature and at high temperature in both inert and reactive gas atmospheres. This suggests that the reduction in abrasion resistance at elevated temperatures is accentuated in the coke with the highest coke reactivity index (CRI) and the highest content of inertinite maceral derived components (IMDC). Our research has shown that at room temperature, the IMDC show greater resistance to abrasion than the reactive maceral derived components (RMDC). This has implications for understanding the relevance of room temperature testing to the abrasion resistance of coke under practical blast furnace operation conditions. For example, a coke from a coal or blend with a comparatively high inertinite content may show lower resistance to abrasion in the blast furnace than would be expected from tumble drum indices. Larger sample numbers and cokes from a variety of single coals need to be tested in further work to verify these assertions. Finally, a novel method using micro-CT image analysis was developed to examine the wear path generated at the coke surface during tribological testing. The purpose of this was to assess the extent of the abrasive wear by comparing the rendered micro-CT images before and after tribological testing. A