Abstract
This dissertation intends to test the effect from a sequence which settled in CM1 / CM2 class. This last one is about French discipline, especially grammatical spelling, and it is going to concern subject-verb agreement. Her aim was to lead student to better manage this type of agreement in a written work context particularly difficult: the student must reconcile several activities simultaneously which may easily cause a cognitive overload. Then the subject-verb agreement mistake are frequently. In order to do this, we have chosen to build our sequence thanks to Allal’s integrated approach (2018): so, every session has taken the form of original loop by mixing both decontextualized grammatical spelling exercises, written work and "Sentence of the day". Seven weeks have been necessary in order to gather our data. Despite the group control absence, the collected data highlight relative progress in some types of learners concerning the third person plural agreement: eventually, the student’s age as well as their initial level would be factor to take into account in order to assess the effect of our sequence.