It is still very challenging to design office buildings to be comfortable in hyper-arid conditions. In this paper, computational fluid dynamics (CFD) has been employed to investigate and improve the thermal performance of an office building in B & eacute;char, Algeria, with ambient temperatures exceeding 40 degrees C. The scenario was analyzed using a complete methodology that integrated field measurements, questionnaires from the occupants, and CFD simulations. The investigation covered two cases: the reference case (Building 1) and a CFD-optimized building envelope (Building 2). The baseline simulation showed that the people were highly dissatisfied with the temperature, with 2.33 PMV and over 65% PPD values for the summer season. The new building envelope, with new insulation and aluminum cladding systems, showed much better improvement in the thermal comfort level. The outcome showed that PMV values were within tolerance (0.5 to +0.5), PPD levels decreased between 30% to 57%, and temperature decreased by about 6 degrees C. High correlation between CFD prediction and field measurement (r = 0.94) shows that the method is reliable. This study proves that CFD is a useful tool to forecast how to design for the climate. It gives evidence-based solutions for keeping individuals more comfortable and using less energy on cooling under weather extremes. The results make a contribution to sustainable building practice in very dry climates and offer a paradigm that can be used repeatedly for improving thermal comfort in poor environmental conditions.