The relationship between perception and production remains an unresolved issue within the study of phonological acquisition. Recent developments in optimality theory offer potentially new solutions to this long-standing problem; but thus far, the proposals that have been advanced are in the absence of actual perception-production data from a given child. This paper provides an empirical instantiation of the perception-production effects by appealing to data from children's knowledge and use of syllables and segments. The demonstration shows that optimality theoretic accounts of the perception-production interface hold in part, but a crucial modification is required in order to obtain the full range of effects that are observed in phonological acquisition. (Contains 1 table and 2 figures.)