The Black population in the United States is becoming increasingly ethnically heterogeneous, making it important to examine ethnicity in the context of mental health.  Prior work has explored within-group variation in mental health conditions among the Black population, but with less attention to African-origin populations, such as Nigerians who account for the largest Black African-origin population in the United States. Using All of Us Research Program data, this study estimates mental health condition prevalence and treatment rates across five Black ethnic groups: African American (n=92,181), Caribbean (n=2,347), Jamaican (n=1,667), Haitian (n=1,195), and Nigerian (n=1,067). We estimated adjusted prevalence of any mental illness (AMI), mood, anxiety, serious mental illness, and neurodevelopmental conditions using logistic regression with three nested models controlling for age, gender, socioeconomic factors, nativity, and insurance, with African Americans as the reference group. Prevalence varied substantially. Nigerian Americans had the highest AMI prevalence (13.1%), with significantly elevated odds relative to African Americans (9.7%), after full adjustment (OR=1.39, 95% CI: 1.15–1.67), driven by anxiety (OR=1.42) and neurodevelopmental conditions (OR=2.82). Haitian Americans had the lowest AMI prevalence (7.9%) and significantly reduced odds in the age- and gender-adjusted model (OR=0.80, 95% CI: 0.65–0.98). Among those with AMI, treatment rates ranged from 55% (Haitian Americans) to 68% (Caribbean Americans); Haitian Americans had significantly lower odds of treatment receipt than African Americans (OR=0.67, 95% CI: 0.45–0.99), with this pattern consistent but attenuating toward the null after full covariate adjustment (OR=0.68, 95% CI: 0.46–1.02). These findings demonstrate that aggregating the Black population in the United States masks clinically meaningful within-group variation and point to the need for targeted outreach, particularly for Haitian Americans facing both lower mental health prevalence recognition and lower treatment engagement.