Children's prosodically conditioned substitutions have been recently described in terms of syllable structure. In this paper we present an alternative analysis, based on the position of the consonant within the foot. We review data from a previous case study (Chiat, 1989) that provide evidence in favour of a foot domain account and against the syllable structure account. One consequence of this finding is that it may be unnecessary to postulate that intervocalic consonants are captured into the coda of the previous syllable. While we caution that more evidence is needed to further test the foot and syllable accounts, we suggest that the foot be considered as a locus of substitution errors in phonology-disordered children.