Objective: To compare the performance of working memory in children of 5 years and 1 month to 6 years and 11 months of public and private Elementary Schools. Methods: 34 subjects were evaluated for both genders composing the GI and GIII public education aged 5 - 6 years; GII and GIV private education aged 5 - 6 years, with an average age of 6 years and 11 months. All subjects were submitted to the Working Memory Assessment Protocol. Results: There were statistically significant differences between the public school and private school in tests of verbal span of serial repetition in direct order with words like phonology and semantics and different words with different semantics and phonology. Between groups, a statistically significant difference occurred between GIII and GIV in verbal span, free recall for polysyllabic words with different semantics and complex phonology; between GI and GII, GIII and GII, GIII and GIV, the verbal span of words with phonology and different semantics, and between GIII and GIV in verbal span of words with the same phonology and semantics. The variables in these groups are age and type of school. Conclusion: With advancing age and grades, the developments of working memory and best performance of working memory were for public school children aged 6 years.