Notwithstanding Mixed OXide (MOX) fuel is considered the most probable option to reduce the separated plutonium stockpiles, in advance on a significant deployment of Fast Breeder Reactors (FBR), inert matrix fuels (IMF) remain a possible way to accomplish this task in nowadays Light Water Reactors (LWR) fleet. A high burning efficiency achieved by preventing new plutonium build-up under irradiation (U-free fuel), a proved high radiation damage and leaching resistance are fundamental requirements when a once-through fuel cycle strategy is planned. Amongst other options, both calcia-stabilised zirconia (csz) and thoria fulfill these criteria standing as the most promising matrices to host plutonium. The investigation of the under-irradiation performance of the calcia-stabilised zirconia fuels is underway, with this regard the thermal conductivity, lower than Uranium OXide (UOX) and MOX fuels, has proved to play a fundamental role. For this reason, ENEA has conceived a comparative in-pile testing (IFA-652 experiment) of viable U-free inert matrix fuel concepts that has been performed in the Halden Heavy Boiling Water Reactor (HBWR) of the Organisation for Economic Co-operation and Development (OECD). The accumulated burn up, under typical LWR conditions, at the end of the irradiation history, was about 90% of the foreseen 45 MWd/kgU(eq). The test-rig is a six-rod bundle loaded with IM, IMT and T innovative fuels. IM and T fuels have, respectively, csz and thoria matrix, the fissile phase is High Enriched Uranium (HEU) oxide (UO2 93% U-235 enriched). I M is a ternary fuel composed of csz+thoria matrix and HEU oxide as fissile phase. Thoria is added in IMT fuel to improve the low IM reactivity feedback coefficients. Pins are instrumented providing fuel centreline temperature, pin inner pressure and fuel stack elongation measurements. 39 This paper is focused on the csz-based IMF under irradiation performance, especially regarding the Fission Gas Release (FGR) topic. Low FGR is fundamental to achieve a high burnup performance and, in the case of IMF, an effective plutonium incineration. Relying on the IFA-652 experimental dataset, especially regarding the ramping-up to the peak power, and the built-in FGR modelling of TRANSURANUS (TU), a well-known nuclear fuel performance code, we expect to provide a contribution on related issues such as the existence of a correlation of IMF FGR onset with the Vitanza threshold and the main processes driving the FGR in these innovative fuels.