the purpose was to analyze the kind of writing mistakes that are modified by the "classwide peer tutoring" (cwpt) program. two second grade elementary school groups participated and were evaluated weekly with a dictation of 20 words in which the number of correct answers and the mistakes according to the academic execution inventory (idea) were measured. the results were that all rule mistakes were reduced while only some specific mistakes were decreased. in addition a difference in the size of the mistakes reduction between both groups was found. the results are discussed in terms of the program's goals and their relationship to the previous literature.