A study re-examined the hypothesis that an identifiable register of child-directed speech (motherese) contributes to child language acquisition. The hypothesis was studied from two perspectives: (1) that it has not been documented adequately at earlier ages; and (2) that individual differences in style of language acquisition interact with maternal measures to mask the effectiveness of motherese. Subjects were 45 mothers and their children at ages 13 months and 20 months. Mother-child interactions were videotaped at each age, and mothers were administered a questionnaire when the child was 13 months to establish the child's language comprehension and production. Two groups of children, earlier and later talkers, were selected for one study, in which maternal language was compared to size of productive vocabulary. Results indicate that the mothers of the groups differed at this stage, suggesting a need for even younger subjects. In the second study, the total sample was divided into two groups according to the child's 20-month stylistic preference (expressive or referential). Lagged associations between maternal 13-month measures and child 20-month mean length of utterance were examined within each group. Results show that maternal variables have different effects depending on the child's adopted strategy, suggesting that in previous research, individual differences have masked the effects of motherese. (MSE)