This paper describes a 2-year longitudinal study of 76 initially prereading children. The study examined the relationships between phonological awareness (measured by tests of onset and rime, phonemic segmentation and phoneme deletion), verbal working memory and the development of reading and spelling. Factor analyses showed that the verbal working memory tests which were administered loaded on two distinct but highly related factors, the first of which,simple repetition, involved the repetition of verbal items exactly as spoken by the experimenter, whereas the second,backwards repetition, involved repetition of items in reverse order. Factor analyses also showed that, whist the phonological awareness variables consistently loaded on the backwards repetition factor at the beginning and end of Grade 1, by Grade 2 the phonological awareness variables loaded on a separate factor which also included sentence repetition. Results of multiple regression analyses, with reading and spelling as a compound criterion variable, indicated that phonological awareness consistently predicted later reading and spelling even when both simple and backwards repetition were controlled. In contrast, verbal working memory did not consistently predict reading and spelling across testing times. Whilst there was some indication that verbal working memory, especially backwards repetition, measured during Grade 1 did predict reading and spelling in Grade 2, these effects were no longer evident when all three phonological variables were controlled. Nevertheless, with 4 individual reading and 2 individual spelling measures as the criterion variables, it was shown that phonological awareness was not quite such a consistent predictor of reading and spelling: it was most highly related to reading pseudowords and spelling real words; but it was not so highly related to spelling pseudowords, apparently because the processing demands of the task for the young children in the study were extremely high. Given the importance of verbal working memory for the completion of phonological awareness, reading and spelling tasks, in particular for spelling pseudowords, the findings are interpreted as providing some support for a theoretical position which posits that both phonological awareness and verbal working memory contribute to the early stages of literacy acquisition. Whilst the findings suggest some support for a general underlying phonological ability, there is also evidence that, as children learn to read and write, verbal working memory and phonological awareness become more differentiated.