Background: Older patients with cancer are at increased risk of hospitalizations during cancer treatment. Our objective was to identify risk factors for unplanned hospitalization among older adults receiving chemotherapy. Methods: We conducted a secondary analysis of a multi-site cohort study (N = 500) of patients ≥ 65 years evaluated with a geriatric assessment (GA) to predict chemotherapy toxicity. The primary outcome wasincident toxicity-related hospitalization during treatment. Predictor variables included: 1) pretreatment GA measures assessing functional status (self-reported and objectively measured), falls, self-reported co-morbidity, mood, social activity and support, nutrition, cognition, and polypharmacy; 2) baseline laboratory values (creatinine clearance [CrCl] by Cockcroft-Gault, hemoglobin, albumin); 3) cancer type and stage; and 4) treatment (mono vs. poly-chemotherapy and dosing). Logistic regression was used to estimate the odd ratio (OR) of hospitalization in multivariate analyses. Results: Among 500 subjects, 117 (23.4%) experienced one or more unplanned hospitalizations during receipt of chemotherapy. Hospitalized subjects were older (74.2 vs. 72.8 years, p = 0.02), more likely to have: a high school or lower education (p = 0.02); a gastrointestinal malignancy (p < 0.001); lower Karnofsky performance status (p = 0.02); polypharmacy ( ≥ 4 medications, p = 0.02); lower instrumental activity of daily living scores (p = 0.03); more comordid conditions (p < 0.001); lower albumin (p < 0.001); and lower CrCl (p = 0.001). In multivariate analyses controlling for cancer type, increased odds of hospitalization was associated with greater number of comorbid conditions (OR = 1.3, 95% Confidence Interval [CI] 1.1-1.5), lower baseline CrCl (OR 1.2, CI 1.1-1.3 per 10 mL/min decrease), and lower baseline albumin (OR 2.7, CI 1.8-4.2 per 1gm/dl decrease), p < 0.001 for all. Conclusions: Readily available information, including self-reported comorbidity, albumin and CrCl, was associated with toxicity-related hospitalizations among older adults receiving chemotherapy. If validated, these data can inform prediction models and interventions designed to decrease hospitalizations.