The work brings results from the five-year research on six groups of students from Charles University. It compares the predictive validity of secondary school performance and entrance examinations.