background: working memory, phonological awareness and spelling hypothesis. aim: to verify the relationship between working memory, phonological awareness and spelling hypothesis in pre-school children and first graders. method: participants of this study were 90 students, belonging to state schools, who presented typical linguistic development. forty students were preschoolers, with the average age of six and 50 students were first graders, with the average age of seven. participants were submitted to an evaluation of the working memory abilities based on the working memory model (baddeley, 2000), involving phonological loop. phonological loop was evaluated using the auditory sequential test, subtest 5 of illinois test of psycholinguistic abilities (itpa), brazilian version (bogossian & santos, 1977), and the meaningless words memory test (kessler, 1997). phonological awareness abilities were investigated using the phonological awareness: instrument of sequential assessment (confias - moojen et al., 2003), involving syllabic and phonemic awareness tasks. writing was characterized according to ferreiro & teberosky (1999). results: preschoolers presented the ability of repeating sequences of 4.80 digits and 4.30 syllables. regarding phonological awareness, the performance in the syllabic level was of 19.68 and in the phonemic level was of 8.58. most of the preschoolers demonstrated to have a pre-syllabic writing hypothesis. first graders repeated, in average, sequences of 5.06 digits and 4.56 syllables. these children presented a phonological awareness of 31.12 in the syllabic level and of 16.18 in the phonemic level, and demonstrated to have an alphabetic writing hypothesis. conclusion: the performance of working memory, phonological awareness and spelling level are inter-related, as well as being related to chronological age, development and scholarity.