Background: Recent studies suggest that ethnic minority students underperform in standardised assessments commonly used to evaluate their progress. This disparity seems to also hold for postgraduate medical students and GP trainees, and may affect the quality of primary health care, which requires an optimally diverse workforce. Aims: To address the following: 1) to determine to what extent ethnic minority GP trainees are more at risk of being assessed as underperforming than their majority peers; 2) to investigate whether established underperformance appears in specific competence areas; and 3) to explore first and second-generation ethnic minority trainees’ deviations. Design & setting: Quantitative retrospective cohort design in Dutch GP specialty training (start years: 2015–2017). Method: In 2020–2021, the authors evaluated files on assessed underperformance of 1700 GP trainees at seven Dutch GP specialty training institutes after excluding five opt-outs and 165 incomplete datasets (17.4% ethnic minority trainees). Underperformance was defined as the occurrence of the following, which was prompted by the training institute: 1) preliminary dropout; 2) extension of the educational pathway; and/or 3) mandatory coaching pathways. Statistics Netherlands (CBS) anonymised the files and added data about ethnic group. Thereafter, the authors performed logistic regression for potential underperformance analysis and χ2 tests for competence area analysis. Results: Ethnic minority GP trainees were more likely to face underperformance assessments than the majority group (odds ratio [OR] 2.41, 95% confidence interval [CI] = 1.67 to 3.49). Underperformance was not significantly nested in particular competence areas. First-generation ethnic minority trainees seemed more at risk than their second-generation peers. Conclusion: Ethnic minority GP trainees seem more at risk of facing educational barriers than the majority group. Additional qualitative research on underlying factors is essential.