ABSTRACTPorphyroblast inclusion fabrics are consistent in style and geometry across three Proterozoic metamorphic field gradients, comprising two pluton‐related gradients in central Arizona and one regional gradient in northern New Mexico. Garnet crystals contain curved ‘sigmoidal’ inclusion trails. In low‐grade chlorite schists, these trails can be correlated directly with matrix crenulations of an older schistosity (S1). The garnet crystals preferentially grew in crenulation hinges, but some late crenulations nucleated on existing garnet porphyroblasts. At higher grade, biotite, staurolite and andalusite porphyroblasts occur in a homogeneous S2 foliation primarily defined by matrix biotite and ilmenite. Biotite porphyroblasts have straight to sigmoidal inclusion trails that also represent the weakly folded S1 schistosity. Staurolite and andalusite contain distinctive inclusion‐rich and inclusion‐poor domains that represent a relict S2 differentiated crenulation cleavage. Together, the inclusion relationships document the progressive development of the S2 fabric through six stages. Garnet and biotite porphyroblasts contain stage 2 or 3 crenulations; staurolite and andalusite generally contain stage 4 crenulations, and the matrix typically contains a homogeneous stage 6 cleavage.The similarity of inclusion relationships across spatially and temporally distinct metamorphic field gradients of widely differing scales suggests a fundamental link between metamorphism and deformation. Three end‐member relationships may be involved: (1) tectonic linkages, where similarP‐T‐time histories and similar bulk compositions combine to produce similar metamorphic and structural signatures; (2) deformation‐controlled linkages, where certain microstructures, particularly crenulation hinges, are favourable environments for the nucleation and/or growth of porphyroblasts; and (3) reaction‐controlled linkages, where metamorphic reactions, particularly dehydration reactions, are associated with an increase in the rate of fabric development. A general model is proposed in which (1) garnet and biotite porphyroblasts preferentially grow in stage 2 or 3 crenulation hinges, and (2) chlorite‐consuming metamorphic reactions lead to pulses in the rate of fabric evolution. The data suggest that fabric development and porphyroblast growth may have been quite rapid, of the order of several hundreds of thousands of years, in these rocks. These microstructures and processes may be characteristic of low‐pressure, firs