Mineralogically, serpentinites consist predominantly of lizardite, clinochrysotile, and antigorite. Recent work has shown that these minerals are not polymorphs. Chrysotile is the only mineral recognized as a synthetic product in experimental studies of the system MgO-SiO₂-H₂O. Antigorite seems to be stable at higher temperatures than lizardite or chrysotile. The density of individual serpentine species is dependent on their morphology; the low-density serpentinites (<2.55g/cc) consist predominantly of clino-chrysotile. Seismic velocities and magnetic susceptibilities of serpentinites are related to the degree of serpentinization. The transition of massive serpentinites from ductile to brittle behavior in laboratory experiments at high confining pressures and temperatures above 300°C has been related to dehydration which may provide a mechanism for developing deep-focus earthquakes along Benioff zones.

Serpentinite is formed by direct hydration of ultramafic protolith in the crust. The most common ultramafic protoliths are harzburgite, dunite, and Iherzolite. The assemblage generally developed from these is lizardite + chrysotile + brucite + magnetite. In areas of high-grade metamorphism, antigorite is the predominant serpentine mineral. The common, large, alpine-type serpentinized ultramafic masses contain brucite and have MgO/SiO₂ ratios similar to those of their protolith, resulting in volume increase during serpentinization. Metamorphic serpentinites and some highly sheared alpine-type serpentinites have lower MgO/SiO₂ ratios than their protolith, lack brucite, and appear t o have been formed by volume-for-volume replacement with concomitant loss of magnesium or addition of silica.

Many large, young masses of peridotite appear to be slabs of oceanic mantle over-thrust onto continental edges. Subsequent sedimentation, serpentinization, and tectonism have greatly modified these original slabs so that their recognition in older orogenic zones is equivocal. The concept of the tectonic evolution of ultramafic rocks from oceanic crust-mantle slabs invading continental margins and being incrementally serpentinized and moved by later tectonic events provides a working hypothesis that allows a better explanation of the many peculiar and varied occurrences of serpentinite. The evidence does not support Hess' suggestion that the third layer of the oceanic crust consists of partly serpentinized mantle peridotite.