Using data from a case study presented in Chiat (1989), Marshall and Chiat (2003) compare two different approaches to account for the realization of intervocalic consonants in child phonology: "coda capture theory" and the "foot domain account". They argue in favour of the latter account. In this note, we present a reappraisal of this argument using the same data. We conclude that acceptance of the foot domain account, in the specific way developed by the authors, is unmotivated for both theoretical and empirical reasons. We maintain that syllable-based coda capture is (still) the better approach to account for the relevant facts. (Contains 3 figures and 8 notes.)