Background: This study evaluated risk factors predicting unplanned 30-day acute service utilization among adult inpatients subsequent to initial hospitalization for a new diagnosis of leukemia, lymphoma, or myeloma, with a focus on social determinants of health. Results from a recent study conducted by the authors indicated that management of medical complications (aligned with OP-35 measure specifications from the CMS Hospital Outpatient Quality Reporting Program) and identification of patients’ psychosocial problems may be targets to prevent unplanned acute care utilization (“readmission”). Methods: The study included 933 unique adult patients admitted to four acute care inpatient facilities within a non-profit community based healthcare system in southern California from 2012 to 2017. Risk models integrated data elements from electronic medical records and facility oncology registries. In addition to univariate analyses, predictive models were constructed using multivariable logistic regression, with standardized coefficients used to rank order variables with greatest impact on readmission. Exploratory data mining techniques were used to augment classification. Results: The mean age of study participants was 65 years; 55.1% were male. Specific diagnoses were lymphoma (48.7%), leukemia (35.2%), myeloma (14.0%) and mixed types (2.1%). The most prevalent insurance type was Medicare/Medicare supplemental plans (49.2%), followed by Medicaid/county services (29.7%). Nearly one-third of patients (31.4%) reported not having a support person or care partner. Approximately one fifth of patients were readmitted. Past history of counseling or use of psychotropic medications was reported by 9.6% of readmitted patients, compared to 5.3% of those who were not readmitted (p=0.03). Readmitted patients were also more likely to report past substance abuse (3.0%) compared to those who were not readmitted (1.1%; p=0.05). Among the risk models, influential social determinants for readmission included lack of consult for social work (lymphoma, myeloma) and past history of counseling or use of psychotropic medications (lymphoma). These factors were not statistically significant in the leukemia risk models. Conclusions: This study provides insights into patient-related factors that may help inform a proactive approach, such as standardized discharge care plans, increased use of social work services, and patient navigation support to improve health outcomes.