Introduction: Several barriers for mental health help-seeking were identified among medical students, including minimizing mental illness. Studies examining aspects particular to those who perceive psychological impairment but do not access treatment are necessary for planning interventions. Aims: To identify help-seeking barriers based on the students’ perception about their need for treatment and psychiatric symptoms. Methods: Cross-sectional study assessing 436 Brazilian medical students (833 attending the medical school = 52.3% response rate). Data collection covered sociodemographic data, mental health, academic environment, and Beck Inventories of Depression (BDI) and Anxiety (BAI). Non-parametric tests and hierarchical logistic regression were used to compare students undergoing treatment, those willing to access treatment, and the ones reporting no need for treatment. Results: Among the 382 students who completed the survey (87.6% completion rate), 38.5% ( n = 147) were in treatment, and 33% ( n = 126) were in need. Moderate to severe risk of alcohol abuse was observed in 45.9% ( n = 50) of students reporting no need for treatment. Regression models suggested that perceiving need for treatment – whether already undergoing it or not – was associated with the severity of depressive symptoms (OR 1.14 [95% CI 1.07–1.21]), female assigned sex (OR 2.18 [95% CI 1.23–3.88]), LGBTQ+ (OR 2.47 [95% CI 1.09–5.60]) and reporting good relationship with the family (OR 0.26 [95% CI 0.08–0.83]). Models comparing students in treatment and those in need, pointed that the factors associated with lacking mental health care were age (OR 0.90 [95% CI 0.82–0.99]), perception of a heavy workload (OR 2.43 [95% CI 1.35–4.38]) and good relationship with colleagues (OR 3.51 [95% CI 1.81–6.81]). Discussion: Social variables and the severity of depressive symptoms are positively associated with perceived need for treatment. Age and academic environment factors were related to help-seeking behavior among students with appropriate self-awareness. We discuss these findings’ implications for planning interventions.