Populations with a migration background have a higher prevalence of mental health problems than their native counterparts. They are also more likely to have unmet medical needs and are less frequently referred to mental health services. One potential explanation for this is that physicians, such as general practitioners (GPs), may unintentionally discriminate against migrant patients, particularly when they lack humanization. To date, no experimental study has investigated this hypothesis. This paper assesses the influence of humanization on GPs' discriminatory decisions regarding migrant patients with depression. A balanced 2 × 2 factorial experiment was carried out with Belgian GPs (N = 797) who received video-vignettes depicting either a native patient or a migrant patient with depression. Half of the respondents were exposed to a text that humanized the patient by providing more details about the patient's life story. Decisions related to diagnosis, treatment and referral were collected, as well as the time spent on each video and text, and were analysed using ANOVA. Migrant patients' symptoms were judged to be less severe than those of native patients (F = 7.71, p < 0.05). For almost all treatments, the decision was less favourable for the migrant patient. Humanization had little effect on medical decisions. We observed that GPs spent significantly more time on the vignette with the humanization intervention, especially for the migrant patients. The results indicate that ethnic differences in the management of depression persist in primary care. Humanization, however, does not mitigate those differences in medical decisions.