Language assessment and evaluation are crucial elements in the language teaching and learning process in particular and Education in general and are considered to be their pedagogical GPS. Yet, their outcome is not always advantageous. The aim of this research is to investigate the negative impact Moroccan Baccalaureate examination has on grammar teaching methodology, learning attitudes, and learning outcomes. The theoretical framework the study starts from is Messick’s (1989) new interactive validity model and Alderson and Wall’s (1993) washback hypothesis. To test its hypothesis, this study employs a mixed-research methodology (MMR), relying both on quantitative and qualitative data collection and analysis instruments, namely, questionnaires comprising yes/no checklists, rating items, Likert scales, and eventually interviews. The results show that (i) most teachers in this study believe that the Baccalaureate examination negatively affects their grammar planning discretion and makes it un-communicative and deductive, (ii) that most students form negative attitudes towards grammar due to the final summative examination, and that (iii) university professors believe that learners come to university linguistically and cognitively unprepared for their academic tasks because of the Baccalaureate negative washback. The research suggests some recommendations to deal with the negative washback impact, namely on the curricular, methodology, and assessment philosophy levels.