If French nursery schools want to play a compensatory role for social inequalities, they have to allow young learners to demonstrate skills (at school) that the most privileged among them can already develop within their family surrounding. These skills account for the best viaticum for their future school curriculums: understanding of written narratives, being able to recount them and acquiring a lexicon. The teaching tool Narramus was created for that purpose. As part of our doctoral research, we wanted to know whether this tool has the expected impact on developing students’ language skills and on teaching methods. For this purpose, we were interested in three questions : -The first one deals with a socio-pedagogical issue: does using an innovative teaching tool make the language learning process of students better and contribute to reduce unequal achievement in learning at school? -The second one deals with a methodological issue: how can we build an experimental set-up in order to evaluate the efficiency of the teaching tool based on strict criteria ? -The third one is related to the role played by the equipment and the support for the improvement of teaching methods: what is the connection between the use of the tool, the support provided by school district teams and the professional development of teachers? In order to answer these questions, we conducted two studies almost experimental within 250 classes, including 1500 students from the first, second and third years of nursery schools. We drew up two experimental protocols abiding by strict methodological criteria defined by the developers of Evidence Based Education. The first study aimed to identify some potential Narramus effect on children learning process and to examine the influence of the support for this innovative tool on the professional development of teachers. Our data showed that students who had access to Narramus had better results than their peers from the control group in all areas assessed. Besides, the differences between the two equipped groups (with or without support) are not significant: the support is not essential to the tool (if well-conceived) to produce the expected positive effects. However, we showed that the support, if it does not directly affect the students’ performances, can although help teachers to identify new professional methods that they can re-use in other learning settings, and thus seems to be a vector of professional development. The second study brought to light the reduction of social inequalities and a transfer of skills built through Narramus to other texts than those studied in class: students who have implemented it understand and recount better than their peers from the same social background and their proficiency level is close to that of more privileged children. Their vocabulary is also enhanced. These conclusions validate the relevance of the tool and contribute to promote an integrated and explicit teaching of comprehension and of vocabulary to young students.