This examination focuses on the idea that child language acquisition is constrained by the same principles that have been found to hold on syllable structure across languages. First, a recently-proposed constraint on syllable structure, the Sonority Cycle, is outlined, and the way that it accounts for syllabic structure across languages is described. Second, it is shown how the Sonority Cycle makes several predictions for child language acquisition when viewed as a principle of universal grammar. Third, data from the literature and from a longitudinal study that bear out the predictions are presented. Focus is on the predominance of consonant-vowel syllables in babbling and early meaningful speech, the imbalance of inventories of syllable-initial consonants as contrasted with syllable-final consonants, and several processes in child speech that have the effect of repairing certain syllable types defined as non-optimal by the Sonority Cycle. It is concluded that the hypothesis that the Sonority Cycle acts as a constraint on the mental representation of syllable structure provides a unified explanation of these data. (MSE)