The widespread Cenozoic Himalayan leucogranites (HLs) are representative rocks evolving from relatively pure crust-derived melts with few mantle material inputs and are believed to be highly fractionated and rare-metal mineralization developed. Over the past several decades, high-quality geochronological and geochemical analyses for the HLs have facilitated exploring the metamorphism, deformation, anatexis, and tectonic settings, enabling us to decipher the evolution process of the Himalayan orogen. In this study, we collected a large amount of geochronological, geochemical, and thermodynamic data from previous studies. Generally, the HLs exhibit (strongly) peraluminous (1.06–1.22) features and are characterized by high SiO2 (70.95–75.08 wt. %) and Al2O3 (14.18–15.85 wt. %) and low TiO2 (0.02–0.26 wt. %), MgO (0.07–0.73 wt. %), MnO (0.01–0.07 wt. %), and total iron (FeOT = 0.44–1.71 wt. %) contents, and are enriched in large ion lithophile elements (e.g. K, Rb, Pb, U). Moreover, rare-metal elements (Li, Be, Ta, W, Sn) also show their enrichments in the HLs; and the deviations of K/Rb (72.16–190.45), Zr/Hf (15.00–35.19), Nb/Ta (2.65–12.75), and Y/Ho (26.07–36.75) ratios from the chondritic values are also notable. All these geochemical structures above suggest that the HLs are (highly) fractionated. Considering the diverging whole-rock isotopic Sr and Nd compositions of the HLs, with the (87Sr/86Sr)t ratios and εNd(t) values mostly ranging from 0.7159 to 0.8052 and −16.78 to −10.43, respectively, the source rocks of the HLs should be multiplex. Additionally, different melting mechanisms (including fluid-absent/present partial melting of muscovite, biotite, and hornblende) can be identified in the HL productions. The occurrences of beryl (Be), columbite (Nb and Ta), spodumene (Li), and many other rare-metal minerals have been frequently reported recently, suggesting the enrichment of rare-metal elements associated with the HLs. To provide a comprehensive and meticulous summary of these HLs (1593 leucogranite samples), we divide them into five stages based on different tectonic backgrounds. After comparing them, we propose that: (1) from Stage I to V, the HL emplacements were strongly related to the tectonic backgrounds: thickened crust from ca. 49 to 40 Ma (I), transition from compression to N-S extension (onset of the South Tibetan Detachment System (STDS)) from ca. 39 to 30 Ma (II), large-scale N-S extension (active movement of the STDS) from ca. 29 to 15 Ma (III), N-S extension to E-W extension tectonic transition (construction of N-S trending rifts (NSTRs)) from ca. 14 to 7 Ma (IV) and rapid uplift of two Himalayan syntaxes from ca. 6 to 2 Ma (V), respectively; (2) the coeval leucogranites emplaced in the Tethyan Himalayas (THLs) and Greater Himalayas (GHLs) had many similarities (particularly in Stages III and IV) in: geochemical compositions, peak emplacement ages, tectonic backgrounds, melting mechanisms, source domains, and fractional crystallization degrees; (3) materials from the Lesser Himalayan Sequence were increasingly significant in producing the HLs from the Eocene to the Pleistocene temporally, and from the Tethyan Himalayan Sequence to the Greater Himalayan Crystalline Complex spatially; (4) owing to the depleted abundances of rare-metal elements in mafic source rocks (e.g. amphibolites) and relatively lower fractional crystallization degrees of the Eocene leucogranites, rare-metal mineralization was rarely reported; (5) excessive fractional crystallization may be the dominant driving force for enriching rare-metal elements (Li-Be-Nb-Ta-W-Sn) of the HLs, making the Himalayas a new potential global orogen scale of polymetallic ore belt. [ABSTRACT FROM AUTHOR]