The article studies whether the use of entry exams or simple admission promote a differentiated academic composition at upper secondary school (USS) in Mexico. The analysis is performed with PISA 2012 data and explores changes in the mathematics performance distribution among lower secondary school students and USS students according to the admission process. I use slope difference analysis, chi-square goodness-of-fit test and quartile regression. The results refute the initial hypotheses since, on the one hand, I observe no increase in the representation of students with better academic results in the states that use entry exams, but rather an under-representation of students with the worse academic results. On the other hand, in the states that use simple admission process I observe the greatest differences in academic performance between lower secondary school students and USS students. Thus, the results show that in both admission processes, the representation of students with the worst performance (1st quintile) decrease significantly in USS and that simple admission does not seem to promote more equal educational transitions for students with different academic performance.