Purpose: Because both t(8;21) and inv(16) disrupt core binding factor (CBF) in acute myeloid leukemia (AML) and confer relatively favorable prognoses, these cytogenetic groups are often treated similarly. Recent studies, however, have shown different gene profiling for the two groups, underscoring potential biologic differences. Therefore, we sought to determine whether these two cytogenetic groups should also be considered separate entities from a clinical standpoint. Patients and Methods: We analyzed 144 consecutive adults with t(8;21) and 168 with inv(16) treated on Cancer and Leukemia Group B front-line studies. We compared pretreatment features, probability of achieving complete remission (CR), overall survival (OS) and cumulative incidence of relapse (CIR) between the two groups. Results: With a median follow-up of 6.4 years, for CBF AML as a whole, the CR rate was 88%, 5-year OS was 50% and CIR was 53%. After adjusting for covariates, patients with t(8;21) had shorter OS (hazard ratio [HR] = 1.5; P = .045) and survival after first relapse (HR = 1.7; P = .009) than patients with inv(16). Unexpectedly, race was an important predictor for t(8;21) AML, in that nonwhites failed induction more often (odds ratio = 5.7; P = .006) and had shorter OS than whites when certain secondary cytogenetic abnormalities were present. In patients with t(8;21) younger than 60 years, type of induction also correlated with relapse risk. For inv(16) AML, secondary cytogenetic abnormalities (especially +22) and male sex predicted better outcome. Conclusion: When the prognostic impact of race, secondary cytogenetic abnormalities, sex, and response to salvage treatment is considered, t(8;21) and inv(16) AMLs seem to be distinct clinical entities and should be stratified and reported separately. © 2005 by American Society of Clinical Oncology.