This paper presents the results of the first phase of the benchmark exercise in the PuMMA project, aimed at extending the reliability of fuel performance codes for Mixed Oxide (MOX) fuels with high plutonium content. The exercise involved the simulation of three irradiation experiments: (1) CAPRIX (45 % of Pu) carried out in the PHENIX reactor in France, (2) TRABANT1 (45 % of Pu) and (3) TRABANT2 (40 % of Pu) irradiated in the HFR reactor in The Netherlands. Thirteen organizations from nine countries participated, using eight different fuel performance codes. In this phase, besides the fuel specifications, boundary conditions such as linear heat rate, irradiation history, and cladding external temperature were provided. In a “blind” exercise, all codes were free to choose suitable models and correlations for the simulation without access to the experimental results. The main results of the “blind” phase are performed along with a sensitivity analysis to investigate the impact of various uncertainties on measurements, irradiation and input parameters. Significant discrepancies were observed among the different codes, attributed to the use of heterogeneous reference correlations and the absence of suitable models for some MOX-specific phenomena. Finally, insights are offered for the second phase of the benchmark, where the fuel performance codes’ capabilities will be assessed against experimental results from Post-Irradiation Examinations of all three pins.