The article addresses the justification of music-structural claims, that is, any statements that assign a structure to a musical passage or a work. It is suggested that a theorist T is justified in making such a claim to the extent that (i) T is able to specify a rule that takes a given musical passage as its input and produces the claim as its output, (ii) T can explain what sorts of facts would render the use of such a rule irrational in musical analysis, and (iii) T has no reason to believe in the existence of such irrationality-conferring facts. Such a view emphasizes the responsibilities of the analyst, but it also gives room for ample methodological pluralism. This is because the facts referred to in (ii) and (iii) may variously be physical, psychological, structural, or doxastic ones. There are thus several conceivable justificatory bases for musical analyses.