This article presents corpus and experimental evidence in support of a parsability-based account of affix ordering in English: an affix that tends to be easily parsed out during speech perception should not occur inside an affix that does not. This generalization holds both at the affix level and the word level. At the affix level, this maxim, when combined with an understanding of the role of frequency and phonotactics in morphological processing, can account for the patterns generally attributed to level ordering. At the word level, it can explain the so-called dual-level behavior of some affixes-an affix may resist attaching to a complex word that is highly decomposable but be acceptable when it attaches to a comparable complex word that favors the direct access route in speech perception. Only a parsing account can afford this set of phenomena a unified explanation.