<p>The transition from high school to university mathematics and the low success rate in the first years of university mathematics is a problem worldwide. One of the most reported factors correlated with academic success is cognitive prerequisites, such as prior knowledge. Aiming to reduce the achievement gap in first-year mathematics courses, we implemented retrieval practice in their lessons: students completed a short test containing two problems at the end of each lesson on the material learned on the given day. The positive effects of retrieval practice – the strategic use of retrieval to enhance memory – have been shown in several cases in university mathematics. Still, the extent to which it can help students of different input levels has been a question. In this paper, we investigated first-year pre-service mathematics teachers’ performance in two mathematics courses: Number Theory and Algebra. We examined students’ test and post-test results relative to their input level using linear regression. The results show that, regardless of the learned mathematical topic, retrieval practice helped minimize the effect of prior academic achievement: students with lower input scores performed similarly to those with higher entry scores on the post-test. In cases where the course was taught without applying the retrieval practice, the achievement gap remained: students with lower entry scores also performed poorly on the course, and those with higher entry scores achieved better results. In retrieval practice groups, the gap between students has narrowed over time, suggesting that students with lower input scores were able to catch up.</p>