Two broad nomenclatures have emerged to describe moult strategies in birds, the 'lifecycle' system which describes moults relative to present-day breeding and other lifehistory events, and the Humphrey-Parkes (H-P) system which reflects the evolution of moults along ancestral lineages. Using either system, challenges have arisen defining strategies in migratory species with more than one moult per year. When all or part of two moults occur in non-breeding areas, extra moults may fail to be recognized or they may have been discriminated temporally, whether feathers are replaced in fall, winter or spring. But in some cases feather replacement can span the non-breeding period, and this has resulted in an inability to identify inserted moults and to compare moult strategies between species. Furthermore, recent analyses on factors influencing the extent of the postjuvenile or preformative moults have either confined this moult to the summer grounds or presumed that it can be suspended and resumed on winter grounds, which has lead to quite divergent results and interpretations. Evolutionarily, the timing, extent and location of moults show phenotypic lability whereas the sequence in which feathers are replaced is comparatively conserved. As, such, I propose defining moults on the basis of feather-replacement sequences as opposed to timing or location of replacement, including strategies in which discrete moults can be suspended for migration and overlap temporally. I provide examples illustrating the functionality of a sequence-based definition in three migratory North American passerines that can undergo feather replacement twice in non breeding areas, and I demonstrate how this system can effectively apply to moults in many other passerine and non-passerine species. I recommend that authors studying the evolutionary drivers of moult strategies in migratory birds adopt a sequence-based approach and to carefully consider replacement strategies both prior to and following autumn migration.