Since Meeussen (1967), there have been recurrent reports of "metatony", a process by which a subset of verb forms ending with L tone utterance-finally end with H tone if followed by an object. While Schadeberg (1986) and Hadermann (2005) speculate that the H is historically derived from the Proto-Bantu *H tone augment and relate metatony to the conjoint/disjoint distinction, it is shown that neither hypothesis accounts for the full range of facts in Abo (A42), a Cameroonian Bantu language closely related to Basaa. Avoiding arbitrary reference to specific TAM forms, the authors present a strictly phonological analysis based on distinct tonal representations.