AbstractRelative alignments of mineral exsolutions and their host crystals can be described by crystallographic orientation relationships (COR). Exsolved phases in garnet from high‐grade metamorphic rocks and igneous rocks may have COR, but the complexity of COR distributions has thus far restricted their use to identifying exsolved phases. Classification of COR also remains mineral‐specific, leaving doubt as to what information COR preserve. To test how COR may be standardized, we calculated mismatch of low‐index crystallographic planes (d‐value ratios) and crystallographic directions (rows of atoms) between precipitates and garnet and defined search criteria for structural alignments likely to be energetically favourable. We analysed published electron backscatter diffraction (EBSD) data for apatite, rutile, ilmenite, corundum, and quartz precipitates in garnet (ntot = 1,296) for the presence of these alignments. Our method predicts between 88% and 98% of observed alignments across the studied minerals and requires only calculations using unit cell parameters. We further show that each exsolved mineral forms COR predicted by the edge‐to‐edge matching model which was developed to describe unambiguous exsolution textures in alloys. Edge‐to‐edge matching aligns atoms at the host–precipitate interface by parallelism or near‐parallelism of crystallographic planes of similar spacing and parallelism of crystallographic directions (rows of atoms) of similar length on the edges of those planes. Edge‐to‐edge matches likely facilitate coherent to semi‐coherent interfaces by lowering surface free energy and strain energy, stabilizing precipitates. These matches are defined by criteria applicable to all minerals, making them an ideal tool for classifying, discovering and interpreting COR of diverse precipitates in garnet. This approach may also predict COR in other geological mineral pairs (e.g., exsolved feldspars). We find that edge‐to‐edge matching may explain the stability of the needle‐shaped morphologies commonly observed for exsolution textures in garnet. Edge‐to‐edge matching COR distributions can be tested as proxies of the state of the host rock at the time of exsolution to evaluate factors such as temperature, cooling rate, degree of undercooling, and/or strain. Patterns of edge‐to‐edge matching COR may be paired with geothermobarometry and/or petrochronology to provide a powerful new tool for studying the histories of granulite and eclogite facies metamorphic and igneous rocks.