This study examines the role of exposure to speech in children's early vocabulary growth. It is generally assumed that individual differences in vocabulary depend, in large part, on variations in learning capacity. However, variations in exposure have not been systematically explored. In this study we characterize vocabulary growth rates for each of 22 children by using data obtained at several time points from 14 to 26 months. We find a substantial relation between individual differences in vocabulary acquisition and variations in the amount that particular mothers speak to their children. The relation between amount of parent speech and vocabulary growth, we argue, reflects parent effects on the child, rather than child-ability effects on the parent or hereditary factors. We also find that gender is an important factor in rate of vocabulary growth. Early childhood is a period of rapid linguistic development. By 2 years, the average child acquires 900 root words (cf. Carey, 1978) and at least a rudimentary syntax (cf. Brown, 1973). Although there has been considerable recent interest in syntactic development, much less attention has been devoted to lexical development. Yet in tracing the development of language from its inception, lexical development must necessarily be a focus of study because the acquisition of words constitutes the child's initial achievement as a language user. A certain amount of vocabulary must be acquired before words can be combined into sentences, and, indeed, several months elapse between the time children start to produce words and the time they start to produce multiword utterances. In addition, vocabulary and syntax are not independent aspects of language knowledge; for example, verbs frequently encode actions involving relations among entities (e.g., give, feed) which are specified by the verb together with its arguments. A major concern in the recent work on syntactic development has been with the relative contributions of the child's innate preparedness for language versus language input. Yet the rapid growth of vocabulary in early childhood also is a manifestation of the human preparedness for language, and parallel questions arise concerning the relative contributions of capacity and input. Especially at the start of language learning, innate preparedness surely plays a role in acquiring word meanings because inferences about meanings are based on pairings of words with situations. As Quine (1969) has persuasively argued, the variety of aspects of a situation which might be encoded by a word is enormous. Because of this, Gleitman and Wanner (1982) point out, it seemscritical to posit innately available constraints on the possible meanings children entertain.