Lunar soil is a fine mixture of local rocks and exotic components. The bulk-rock chemical composition of the newly returned Chang’E-5 (CE-5) lunar soil was studied to understand its chemical homogeneity, exotic additions, and origin of landing site basalts. Concentrations of 48 major and trace elements, including many low-concentration volatile and siderophile elements, of two batches of the scooped CE-5 soil samples were simultaneously obtained by inductively coupled plasma mass spectrometry (ICP-MS) with minimal sample consumption. Their major and trace elemental compositions (except for Ni) are uniform at milligram levels (2–4 mg), matching measured compositions of basaltic glasses and estimates based on mineral modal abundances of basaltic fragments. This result indicates that the exotic highland and KREEP (K, rare earth elements, and P-rich) materials are very low (<5%) and the bulk chemical composition (except for Ni) of the CE-5 soil can be used to represent the underlying mare basalt. The elevated Ni concentrations reflect the addition of about 1 wt% meteoritic materials, which would not influence the other bulk composition except for some highly siderophile trace elements such as Ir. The CE-5 soil, which is overall the same as the underlying basalt in composition, displays low Mg# (34), high FeO (22.7 wt%), intermediate TiO2 (5.12 wt%), and high Th (5.14 µg/g) concentrations. The composition is distinct from basalts and soils returned by the Apollo and Luna missions, however, the depletion of volatile or siderophile elements such as K, Rb, Mo, and W in their mantle sources is comparable. The incompatible lithophile trace element concentrations (e.g., Ba, Rb, Th, U, Nb, Ta, Zr, Hf, and REE) of the CE-5 basalts are moderately high and their pattern mimics high-K KREEP. The pattern of these trace elements with K, Th, U, Nb, and Ta anomalies of the CE-5 basalts cannot be explained by the partial melting and crystallization of olivine, pyroxene, and plagioclase. Thus, the mantle source of the CE-5 landing site mare basalt could have contained KREEP components, likely as trapped interstitial melts. To reconcile these observations with the initial unradiogenic Sr and radiogenic Nd isotopic compositions of the CE-5 basalts, clinopyroxene characterized by low Rb/Sr and high Sm/Nd ratios could be one of the main minerals in the KREEP-bearing mantle source. Consequently, we propose that the CE-5 landing site mare basalts very likely originated from partial melting of a shallow and clinopyroxene-rich (relative to olivine and orthopyroxene) upper mantle cumulate with a small fraction (about 1–1.5 %) of KREEP-like materials.