There are insufficient high-grade iron ores currently being mined to meet world demand for steel production. In order to meet raw material demand in India, lower grade ores with high alumina contents are being crushed and beneficiated, mainly by gravity techniques requiring water. However, the scarcity of water in the mining areas warrants the development of some dry gradation techniques for iron ores so that the inferior ore specimens can be rejected in order to improve the grade of the concentrate. The present gradation of ores by mineralogical/chemical methods is time-consuming and cumbersome. This paper presents an Infrared (IR) thermography-based non-invasive technique for the faster gradation of iron ores, taking into account the variation in thermal absorptivity of the ore constituents. Iron ore samples from the Joda, Noamundi and Barsua mines, with Fe contents ranging from 52 to 67 wt.%, were selected and crushed to around 10 mm particle size. The crushed iron ores were uniformly heated using a microwave oven, for a time period sufficient to create a difference between the ore particles in the extent of their respective infrared emissions. The thermal images of the heated particles were captured by IR thermography and the peak temperature of each ore particle was obtained from the thermal profile. A computer program was developed for ore gradation based on the peak temperature of each ore particle which corresponds to its iron content. A threshold was selected through chemical verification of the ores and accordingly the ores were categorized as high-, medium- and low-grade.