The limited access of low social class pupils to higher education is a pervasive phenomenon that contributes to the reproduction of social inequalities. We propose to investigate structural obstacles that may contribute to actively exclude low social class pupils from higher education, by focusing on the use by evaluators of merit-based selection practices (grading and tracking). We test the hypothesis that selection practices can be biased by the evaluators’ awareness of the pupils’ socioeconomic status (SES). The first line of research (Autin et al., 2019) tested the hypothesis that normative assessment (i.e., grades), compared to formative assessment (i.e., comments), produces performance differences that reflect social inequalities. Participants had to assess a test supposedly produced by a low vs. high SES pupil using a normative vs. a formative method. Participants detected more mistakes in the test of a low-SES pupil than in that of a high-SES pupil, only when they used a normative assessment. The second line of research (Batruch et al., 2019) focused on school tracking. Participants had to decide which secondary school track was most suitable for a 12 year-old student. Every target student was presented with the same grades, which were slightly below the usual requirements for the highest track, and the student’s SES was manipulated. Participants considered the lower track more suitable for a low-SES student than for a high-SES student, while the higher track was deemed more suitable for a high-SES student. These findings suggest that selection practices reinforce social class inequalities by allowing the agents of the educational system to actively constrain the access to educational attainment (Butera et al., 2021).