The assessment of vocabulary knowledge has posed a number of problems which have resulted in widely differing estimates of children���s vocabulary growth. Such estimates assume importance when school texts are controlled for vocabulary diversity, and in pedagogical debates concerning the value of vocabulary instruction. In this paper we outline a computer���based methodology which obtains simple timed responses to sets of words and word���like nonsense items. We present pilot data depicting estimated vocabulary knowledge, and relating it to word frequency and decision response time, and we speculate about the question of total vocabulary size. �� 1985 Taylor & Francis Group, LLC.