Purpose Prevalence rates, and bivariate comorbidity patterns, of many common mental disorders differ significantly across ethnic groups. While studies have examined multivariate comorbidity patterns by gender and age, no studies to our knowledge have examined such patterns by ethnicity. Such an investigation could aid in understanding the nature of ethnicity-related health disparities in mental health and is timely given the likely implementation of multivariate comorbidity structures (i.e., internalizing and externalizing) to frame key parts of DSM-5. Methods We investigated whether multivariate comorbidity of 11 common mental disorders, and their associated latent comorbidity factors, differed across five ethnic groups in a large, nationally representative sample (n = 43,093). We conducted confirmatory factor analyses and factorial invariance analyses in White (n = 24,507), Hispanic/ Latino (n = 8,308), Black (n = 8,245), Asian/Pacific Islander (n = 1,332), and American Indian/Alaska Native (n = 701) individuals. Results Results supported a two-factor internalizing- externalizing comorbidity factor model in both lifetime and 12-month diagnoses. This structure was invariant across ethnicity, but factor means differed significantly across ethnic groups. Conclusions These findings, taken together, indicated that observed prevalence rate differences between ethnic groups reflect ethnic differences in latent internalizing and externalizing factor means. We discuss implications for classification (DSM-5 and ICD-11 meta-structure), health disparities research, and treatment. Springer-Verlag (outside the USA) 2012.