Because all high-silica granitic rocks (SiO₂>70%) are dominated by quartz and feldspars, they have broadly similar major element compositions. Despite the low abundances of the oxides FeO, MgO, CaO, Na₂O, and K₂O that form the basis of the geochemical classification of granites (Frost et al., 2001), it is possible to identify six geochemical subgroups. These include:

1. Peraluminous leucogranites (Harney Peak granite, South Dakota). The peraluminous leucogranites may be alkalic to calcic, and magnesian to ferroan. They are characterized by low Zr, typically less than 50 ppm, reflecting their low temperature of formation by partial melting of pelitic rocks.
2. Calc-alkalic metaluminous leucogranites (Tuolumne intrusion, Sierra Nevada batholith). These leucogranites are magnesian to ferroan. Zr contents are typically less than 100 ppm. They are interpreted as late-stage differentiates of calc-alkalic granitic magmas.
3. Trondhjemites (Fagervika, Trondheim, Norway). These are calcic, magnesian, mainly peraluminous leucogranites that contain moderate amounts of Zr (100-150 ppm). They are considered to form by partial melting of amphibolite.
4. Ferroan calc-alkalic peraluminous granites (Jamon and Musa granites, Amazonia). These granites have Zr contents of 150-400 ppm and form by partial melting of tonalite or granodiorite crust.
5. Calcic ferroan leucogranites (Webb Canyon Gneiss, Teton Range). These leucogranites are markedly calcic, and metaluminous to peraluminous. Zr varies from 200-900 ppm. Their higher Zr content reflects formation at higher temperature than peraluminous leucogranites.
6. Peralkaline ferroan leucogranites (Brandberg, Namibia). These peralkaline granites are strongly ferroan, and alkali-calcic to alkalic. They represent the last stage of differentiation of peralkaline melts. They may have extreme Zr contents (in excess of 1000 ppm), reflecting the complexation of Na with Zr in the melt, suppressing zircon crystallization.

Examination of the compositions of these high-silica granites indicates that their differences arise from variables including the composition of the source rock, the temperature of partial melting, and the role of fractional crystallization, and confirms that there is no single solution to the “granite problem.”