Introduction: Accountable Care Organizations (ACOs) aim to improve quality and reduce costs of care, but few studies have described variation across ACOs in hospitalization rates for patients with heart failure (HF) or factors associated with hospitalization rates.Methods: We identified a sample of Medicare fee-for-service beneficiaries with HF who were assigned to a Medicare Shared Savings Program (MSSP) ACO in 2017 and survived at least 30 days into 2018. Using a hierarchical, negative binomial model that accounted for clustering of patients within ACOs, we calculated 2018 risk-standardized, unplanned ACO admission rates (RSAARs) as the ratio of predicted to expected admissions per 100 persons, multiplied by the overall rate of admissions. We then used multiple linear regression to identify ACO characteristics associated with RSAAR variation.Results: Among 1,232,222 beneficiaries with HF, 283,795 were assigned to one of 467 MSSP ACOs (mean age 81 years, 54% female, 86% white, 78% urban). Median RSAAR [IQR] was 87 [82-92] admissions per 100 persons (Figure 1). A 5% increase in the percentage of Black beneficiaries in the ACO corresponded to an increase of 0.65 admissions per 100 HF patients (95% CI 0.31, 0.99, p<0.001). ACOs in the Northeast had higher RSAAR (B=5.16, 95% CI 3.25, 7.07, p<0.001), while ACOs in the West had lower RSAAR (B=-4.36, 95% CI -6.64, -2.09, p<0.001) (Table 1).Conclusions: Among Medicare beneficiaries with HF, there is variation in admission rates across ACOs. Region was strongly associated with RSAAR. Further studies should examine whether regional variation in admission rates is related to ACO quality or regional variation in admission more generally.