The objective of this work is to study the influence of sleep temporal allocation at the nights before the assessments on the academic performance in medical students at the beginning of the course at UFRN - Central Campus. This is a cross-sectional study conducted in the semesters 2018.2, 2019.1 and 2019.2. Students filled out an identification form, answered the Morningness-Eveningness Questionnaire (MEQ), Munich ChronoType Questionnaire (MCTQ), Pittsburgh Sleep Quality Index (PSQI), Beck Depression Inventory (BDI) and the State-Trait Anxiety Inventory (STAI). For academic performance, the final grade of the discipline Biological Module II was used for students in the second semester and the final grade of the discipline Infectology for students in the fourth semester. The Kruskal Wallis test was used to assess the relationship of sleep temporal allocation at the nights before the assessments with academic performance, Horne and Östberg chronotype score, Social Jet Lag (SJL), PSQI score, BDI score, anxiety-state and trait anxiety scores. The research was approved by the UFRN Research Ethics Committee with opinion number: 2,996,461. Our results showed that the majority of the 201 students that participated in the study are male (62.2%), single (94.5%), with average age, body mass index (BMI) and academic performance of 21, 45 ± 3.47 years, 23.66 ± 3.67 and 7.41 ± 0.96 values respectively. Only 30.3% of students reported that they did not change their sleep routine at the nights before the assessments. The majority, 48.8% reported that they use the behavioral strategy of staying up late and sleeping a little before the assessments. Around 5% reported that they do not sleep before the assessments. We found a statistically significant difference between the groups according to the sleep temporal allocation at the nights before the assessments with academic performance [KW = 15.612 (4); p = 0.004], Horne and Östberg chronotype score [KW = 36.936 (4); p <0.01], PSQI score [KW = 22.213 (4); p <0.01], BDI score [KW = 19.358 (4); p = 0.001], anxiety-state score [KW = 23.316 (4); p <0.01] and trait anxiety score [KW = 16.883 (4); p = 0.002]. These results lead us to conclude that students who use a strategy of sleep temporal allocation at the night before the assessments that causes total or partial sleep deprivation have worse academic performance, tendency to evening-type, poor sleep quality and high scores of depression and anxiety. These findings reinforce the guidelines for students to have a study routine. In addition, it has a pedagogical implication in the sense of the need for the teacher to have a mechanism to follow the study routine and thus guide students.