Compared with widespread Egyptian Tertiary basaltic rocks, Mir Tertiary basaltic dykes from the Western Desert of Egypt show distinct mineralogical and geochemical characteristics that provide insights into the nature of their mantle source and geotectonic evolution as well as melting and crystallization conditions. Their mineralogical features (i.e., presence of orthopyroxene and lack of olivine) distinguish them from the other Egyptian Tertiary basalts. They have basaltic andesite compositions (SiO2 = 53.55–55 wt.%, Na2O + K2O = 4.64–4.8 wt.%, MgO = 2.66–3.02 wt.%) with tholeiitic character. Temperature of crystallization ranges from 1,112°C to 1,140°C for plagioclase; from 1,085°C to 1,123°C for clinopyroxene; and from 1,034°C to 1,090°C for orthopyroxene. These results are in harmony with petrographic observations which suggest early crystallization of plagioclase, followed by simultaneous crystallization of plagioclase and clinopyroxene, and soon thereafter by orthopyroxene crystallization. Titanomagnetite–ilmenite intergrowths indicate oxygen fugacity (ƒO2) of −12.09 and −9.4 log units, suggesting crystallization at relatively oxidized conditions. Their temperature ranges from 959°C to 1,125°C, suggesting that the titanomagnetite joins the crystallized phases at early stages and continued until later stages. Mir basaltic andesite (MBA) has low Ni (15–27.9 ppm), Cr (19.8–48.5 ppm), and Sc (20–25 ppm) concentrations together with high FeOt/MgO ratios (3.18–3.58) relative to primitive basalts, arguing for their evolved nature. They have sub‐chondritic Nb/Ta ratios (15.5–17) similar to ocean–island basalts and higher than those of continental crust, ruling out a significant effect of crustal contamination. The high La/Nb (>1.4) with the negative Nb, Ta, and Ti anomalies on the primitive mantle‐normalized patterns most likely indicates that their primary magma was derived from partial melting of a mantle source previously metasomatized by influx of subduction‐related slab fluids/melts. Moreover, the super‐chondritic Zr/Hf ratios (39.55–55.21) and the negative K and Ti anomalies point to carbonate metasomatism of their mantle source. This inference is further supported by their high (K2O + Na2O)/TiO2 ratios (>1) which are comparable with experimentally produced carbonated peridotite melts. The high Zr/Y (average 9.75) and La/Yb ratios (8.6–10) of the MBA suggest that they were generated via low degrees of partial melting. Indeed, modelling calculations indicate generation by lower melting degrees (<10%) of a garnet‐bearing lherzolite mantle source. The melting region was at the garnet stability field suggesting a deeper origin (exceeds ~85 km) for MBA. Field and geochemical characteristics of the Egyptian Tertiary volcanism proved that it is spatially and temporally related to extension that has occurred during Red Sea rifting. Extension most likely due to the NE–SW tensional stresses which reactivated the old fractures and triggered mafic magma generation through passive upwelling of the asthenospheric mantle.