This contribution compares progressive deformation of granite and low-grade metasedimentary rocks near the brittle-ductile transition in the strike-slip Kellyland fault zone (KFZ). Fault rocks from both protoliths preserve textures indicative of diffusion-assisted granular flow, but they record different strain-localization mechanisms and rheologies. The SE side of the KFZ preserves a transition from pulverized granite through cataclasite into granite-derived ultramylonite. Textural relationships here indicate that ultramylonite formed after brittle grain-size reduction during seismogenic faulting. This interpretation is supported by abundant cm-scale mylonite/ultramylonite zones formed along brittle fractures cutting pulverized granite and the compositional similarity of ultramylonite, mylonite, and undeformed granite. Pelitic metasedimentary rocks cut by the KFZ are texturally and mineralogically very similar to pelitic rocks outside the zone and therefore likely underwent little deformation weakening. Wacke intervals cut by the KFZ preserve evidence for conversion of silt and sand-sized grains to fine-grained matrix via reaction weakening. Metasedimentary rocks cut by the KFZ also host variably deformed quartz +/- calcite veins, and pressure-shadow overgrowths on porphyroclasts are ubiquitous. Granite-derived mylonitic rocks exhibit very little mineral deposition in dilational sites, however. Quartz recrystallized grain sizes indicate that granite-derived ultramylonites supported 2.1–2.5-times more stress than texturally similar metasedimentary rocks in the fault core. The variable strain-localization mechanisms across the KFZ highlight the importance of protolith in controlling the local evolution of crustal-scale faults. The variable evidence for elevated pore-fluid pressure and fluid-enhanced syndeformational reactions indicates that across-strike fluid flow was restricted. The variable stress record across the planar KFZ indicates that granite and metasedimentary rocks record different deformation phases. One explanation for this is a period of fault slip where granite existed on both sides of the KFZ in the study area. The Amazon Mountain pluton exposed ~4 km NE of the study area is a likely culprit, but outcrop is insufficient to test this hypothesis.