Like active labour market programmes, grade repetition could generate two types of effects: better/worse outcomes due to programme participation (i.e. the fact that pupils repeat a particular grade). This is what the existing literature on grade repetition has focused on. Another potential outcome is the "threat" effect of grade repetition. Pupils and/or their family could make significant efforts to avoid grade repetition and its important opportunity cost. Learning effort by pupils could be a function of the risk of grade repetition. This paper attempts to assess that relationship by exploiting a reform introduced in 2001 in the French-Speaking Community of Belgium, synonymous with a reinforced overall threat of grade repetition. The possibility to impose grade repetition sanctions and the end of grades 8-12 has always existed, but in year 2001, policy makers reinstated the possibility to repeat grade 7, putting an end to the regime of "social promotion" applicable to that grade since 1995. We use data from two waves of the Programme for International Student Assessment study (corresponding to periods before and after the reform) to evaluate the medium-term effects of this reform. The first measure of performance we consider is the position in the curriculum (or grade) reached at the age of 15, and we show that it deteriorated after 2001. We also consider the reform's impact on test scores. Focusing on grade 10, we fail to verify the necessary condition for grade repetition threat to lead to higher test scores. The tentative conclusion is that an enhanced threat of grade retention after 2001 did not lead to better medium-term outcomes, even among the segments of the population the most at risk of grade repetition.