The present study aimed to identify that For the conduct of the research, the quantitative approach was chosen, in which secondary data of all students enrolled between the years 2017 and 2019 were used. Descriptive and inferential methods were used to assist the analyzes. It was found that of all the variables analyzed, six presented some type of association with the evasion: age of entry, time between high school and inclusion in higher education, school performance coefficient (CRE), failure, amount of periods studied and income, with emphasis on category "undeclared". The research relevant to the analyzed institution, because, evidenced that the highest proportion of evasion is occurring in the initial periods, among students with very low CRE, high failure for lack, high age and that students with undeclared income are not attending the classes.