Background: Unanticipated healthcare resource utilization, in the form of either emergency department utilization (EDUs) or hospital admission (HA), is an indicator of lower quality of cancer care. The objective of this study was to develop a comprehensive predictive model for emergency department utilization and hospital admission (HA) for patients with solid tumors who received chemotherapy within 30 days. Methods: We abstracted electronic health data on oncology encounters from all patients receiving systemic therapy chemotherapy for solid tumors from March 01, 2015 to August 21, 2020. Patients included in the cohort met the following criteria:1) Ages > = 18 years old 2) Diagnosed with any solid tumor (3) Had any oncology treatment plan related to their solid tumor diagnosis; 4) Received at least one chemotherapy during March 01, 2015 through August 21, 2020. We defined a primary composite outcome of emergency visit or unanticipated hospitalization within 30 days after the encounter. Patient information was abstracted. We developed a predictive model for the primary outcome using LASSO logistic regression modeling. To evaluate the model, we calculated the Area Under the Receiver Operator Characteristic (AUROC) and the calibration slope. Results: 12,917 unique patients with 134,641 encounters were included. Of these, 9100 (6.7%) of encounters were ED visits or hospitalizations. 714 patients (5.5%) had at least one ED visit or hospitalization within 30 days of treatment; patients at highest risk for this outcome included those with pancreatic cancer, lung cancer and ovarian/endometrial cancer (9.1%, 7.4% and 6.8% respectively). 38 variables were identified as contributing to the LASSO model. The top predictors, in order of importance, were fever, hypernatremia, hypotension and low albumin. The model’s AUC was 0.83, indicating good model sensitivity to outcome. Conclusions: Approximately 7% of healthcare encounters in patients undergoing treatment for solid malignancies were ED visits or hospitalizations. Identifying patients at highest risk for this outcome may prevent unnecessary resource utilization and optimize patient care during cancer treatment. The model developed in this study demonstrated good sensitivity and could be implemented in clinical practice to allow for preventive outpatient interventions. Work is ongoing to integrate this model into clinical care at our institution.