Petrofabric contributions are here classified into seven categories, namely: (1) descriptive techniques; (2) kinematic analyses; (3) geometric analyses of folds; (4) absolute strain analyses; (5) studies of experimentally deformed single crystals and rocks; (6) dynamic analyses of deformation features in naturally deformed rocks; and (7) studies for engineering rock mechanics. The nature of each subfield is given, and items (2) and (6) are discussed in detail.The principles of kinematic analysis through symmetry are reviewed with the aid of simple models of brittle and ductile deformations in which componental displacements are specified and the symmetry of the initial and final fabric is known. The models illustrate the degree of similarity among the symmetries of the fabric, componental displacements, deformation (strain and rigid-body rotation), and states of stress that exist in rocks at the time a fabric element forms. Both isotropic and anisotropic models are considered. In the more realistic cases the symmetry of the pattern of componental displacements is identical to the symmetry of the resulting fabric. If the body behaves isotropically during the deformation, symmetry elements in the final fabric will be common to corresponding elements in the symmetries of the states of stress and deformation. If, however, the body behaves anisotropically one cannot relate symmetry elements in the final fabric to the symmetry of the states of stress and deformation without making assumptions about the symmetry of the anisotropy. The study of fabric symmetry may contribute some detail to the geometric pattern of the strain or displacement fields, but is otherwise of little use in furthering understanding of the mechanics of geologic deformation.Results of dynamic analysis, which involve determination of the orientations and relative magnitudes of principal stresses in naturally deformed rocks, are critically reviewed. The principal axes inferred from fractures, twin lamellae, deformation lamellae, etc., are those within the material at the instant the fabric element forms. As such they correspond to the incremental stresses which are coincident with the incremental strains in linear materials. Certain generalizations are drawn from previous work on slightly and moderately deformed rocks in which the orientation of the incremental stresses and strains are probably coincident with the total stresses and strains. The critique leads to a working philosophy to guide future studies, namely: dynamic petrofabric analyses provide the means for testing hypotheses generated through theoretical and experimental studies of rock deformation. A study should be designed to test specific hypotheses. Samples should be collected from critical domains where the results may provide a basis of choice between alternative hypotheses. The techniques need to be refined to distinguish small differences in the orientations of principal stress axes deduced from fabric data. The chronological sequence in the development of the fabric elements requires attention. Several suggestions along these lines are offered.