Introduction: Heart failure (HF) readmission quality metrics disproportionately impact reimbursement in safety net hospitals. Prior research has demonstrated the effect of medical comorbidities on readmission, however, there is a paucity of data on predictors of readmission in vulnerable and underserved HF patients. We sought to evaluate the effect of demographics, medical and social comorbidities on risk of 30 day readmission in an academic safety net hospital in San Francisco. Methods: We performed a retrospective chart review from 2018 to 2020. Patients were included if treated for HF while on inpatient cardiology or medicine services and were assigned an ICD-10 discharge code for HF. Patients less than 21 years old were excluded. Demographics and comorbidities were obtained through evaluation of ICD-10 discharge codes and chart review. Multivariate modeling was used to determine predictors of 30 day readmission. Results: The study population included 383 patients in which the mean age was 60±13 years and 73% (n=282) were male. 44% (170) were Black, 23% (88) were Latinx, 33% (127) were not housed, 97% (371) had public insurance, and 21% (81) had a diagnosis of mental illness. 46% (177) had CAD, 76% (291) hypertension, and 36% (177) DM. Substance use was common with 30% (114) using methamphetamines, 36% (138) cocaine, 18% (69) opioids, and 35% (135) alcohol. On multi-variate analysis, EF less than 40% (75%, 285) was the only medical comorbidity associated with an increased risk of readmission (OR 1.86, 1.1-3.1, p= 0.018). Social variables associated with increased risk of readmission included identifying as Black (OR 2.26, 1.03-5.0, p= 0.043) or Latinx (OR 3.43, 1.41-7.59, p= 0.006), homelessness (OR 3.02, 1.76-5.18, p=<0.001), and specific substance use: methamphetamine (OR 2.23, 1.39-3.57, p=0.001), cocaine (OR 1.63, 1.03-2.57, p= 0.037), opioids (OR 1.81, 1.05-3.13, p= 0.033), and alcohol (OR 2.26, 1.43-3.58, p= 0.001). Conclusion: Race, housing status and substance use were more strongly associated with readmission risk than medical comorbidities in a population of urban, vulnerable and underserved HF patients. Interventions to improve HF readmission metrics should consider addressing racial and social disparities in similar populations.