Analyses of 230 Franciscan rock and mineral samples, including the San Luis Obispo ophiolite, show that metamorphism produces no change in the δ 18O of the graywackes (+11 to +14), but that igneous rocks become enriched in 18O by 2–6% and the cherts depleted by 5–10%. The shales are of two types, a high- 18O type (+16 to +20) associated with chert and a low- 18O type isotopically and mineralogically similar to the graywackes. The vein quartz ( δ = + 15 to + 20) is invariably richer in 18O than the host rock quartz and in most of the rocks the δ 18O of the clastic quartz is similar to the δ 18O of the whole rock. Mineral assemblages are typically not in isotopic equilibrium. Although the δ 18O values are very uniform (+13 to +16). the δ 13C of vein aragonite and calcite is widely variable (0 to − 14), implying that a major source of the carbon is oxidized organic material. The δD values of 83 igneous and sedimentary rocks are -45 to -80, exceptions are the Fe-rich minerals howieite and deerite, which have δD = −100. All of these samples could have equilibrated with H 2O having δD ≈ +10 to −20 and δ 18 O ≈ −3 to +8. assuming temperatures of 100–300°C. However, the serpentines ( δD ≈ −85 to −110) and the vein minerals ( δD = −23 to −55) are exceptions. The vein minerals are 10–20%, richer in deuterium than the adjacent wall rocks; they formed from a relatively D-rich metamorphic water, typically at lower temperatures than did their host rocks. The isotopic compositions of the other Franciscan rocks were affected by three distinct events: (1) hydrothermal alteration of the ophiolite complexes and volcanic rocks as a result of submarine igneous activity at a spreading center or in an island-arc environment; (2) low-temperature, high-pressure regional metamorphism and diagenesis; and (3) a late-stage, very low temperature (<100°C) alteration of the ultramafic bodies by meteoric ground waters, producing lizardite-chrysotile serpentine. In the first two cases, the pore fluid involved in the alteration of the Franciscan rocks was sea water. However, this water became somewhat depleted in D and enriched in 18O during blueschist metamorphism, evolving to values of δD ≈ − 20 and δ 18 O ≈ + 6 to + 8 at the highest grades. Except for one graywacke sample, the meteoric waters that affected the serpentinites did not significantly change the D H ratios of the OH-bearing minerals in any other Franciscan rock. The δ 18O values of orogenic andesites are too low for such magmas to have formed by direct partial melting of Franciscan-type materials in a subduction zone. Andesites either form in some other fashion, or the melts must undergo thorough isotopic exchange with the upper mantle. The great Cordilleran granodiorite-tonalite batholiths, however, are much richer in 18O and may well have formed by large-scale melting or assimilation of Franciscan-type rocks. The range of δD values of Franciscantype rocks is identical to the −50 to −80 range shown by most igneous rocks. This suggests that ‘primary magmatic H 2O’ throughout the world may be derived mainly by partial melting of Franciscantype materials, or by dehydration of such rocks in the deeper parts of a Benioff zone.