Post-stroke depression (PSD) has been proven to be associated with stroke severity. Thus, we hypothesized that the prevalence of PSD would be lower in patients with mild stroke. We aim to explore predictors of depression at 3 months after mild acute ischemic stroke (MAIS) onset and to develop a practical and convenient prediction model for the early identification of patients at high risk. A total of 519 patients with MAIS were consecutively recruited from three hospitals in Wuhan city, Hubei province. MAIS was defined as a National Institute of Health Stroke Scale (NIHSS) score of ≤5 at admission. Meeting the DSM-V diagnostic criteria and a 17-item Hamilton Rating Scale for Depression (HAMD-17) score of >7 at their 3-month follow-up were considered the primary outcomes. A multivariable logistic regression model was used to determine the factors adjusted for potential confounders, and all independent predictors were brought into the construction of a nomogram to predict PSD. The prevalence of PSD is up to 32% at 3 months after MAIS onset. After adjusting for potential confounders, indirect bilirubin (p = 0.029), physical activity (p = 0.001), smoking (p = 0.025), hospitalization days (p = 0.014), neuroticism (p < 0.001), and MMSE (p < 0.001) remained independently and significantly related with PSD. The concordance index (C-index) of the nomogram jointly constructed by the aforementioned six factors was 0.723 (95% CI: 0.678-0.768). The prevalence of PSD seems equally high even if the ischemic stroke is mild, which calls for great concern from clinicians. In addition, our study found that a higher level of indirect bilirubin can lower the risk of PSD. This finding may provide a potential new approach to PSD treatment. Furthermore, the nomogram including bilirubin is convenient and practical to predict PSD after MAIS onset.