High Ba-Sr granites are geochemically distinct from the more familiar I-, S-and Atypes (of igneous, sedimentary or anorogenic/anhydrous/alkaline parentage), in particular by their lack of depletion in Ba and Sr relative to other large-ion lithophile elements (LILE). These differences are sufficient enough to indicate different petrogenetic processes which are still the subject of considerable debate. High Ba-Sr plutons from the classic Caledonian type location in NW Scotland show a range of elemental and isotopic enrichment, ideal for novel petrogenetic investigations such as the application of "non-traditional" stable isotope systems. Two directly comparable plutons have been analysed for Ba and Mg isotopes, from either end of the range of enrichment. Barium isotopes (δ 138/134 Ba values of -0.05 to +0.23‰ in the relatively "depleted" Strontian pluton and -0.19 to +0.03‰ in the enriched Rogart comparator) overlap normal mantle values but extend to lighter compositions, consistent with addition of up to 5% pelagic sediment that has incorporated variable amounts of biogenic barite. Magnesium isotopes (δ 26 Mg values of -0.30 to -0.25‰ versus -0.42 to -0.23‰ in Strontian and Rogart, respectively) lie largely within the normal mantle range, with two light outliers possibly indicative of carbonate involvement but for which fractional crystallization cannot be excluded. Associated relationships between fluid-mobile and melt-mobile elements suggest that sediment melting transfers the crustal isotope signature to a dominantly mantle source, itself variable between isotopically depleted mantle wedge and previously enriched lithospheric mantle. The combined Ba, Mg, O, Sr, and Nd isotope and elemental data are all consistent with a small percentage of pelagic sediment in a mantle source for high Ba-Sr granites, and this is sufficient to disguise these essentially mantle-derived rocks as reworked continental crust. First-order estimates suggest that some 10% of granites worldwide may have high Ba-Sr character, although the proportion of these that represent largely juvenile origin is currently unknown. Nevertheless, as compositional equivalents of Archean sanukitoids, they may represent unrecognized mantle contributions to crustal growth over some 3 billion years of Earth history.