Background: Presence and severity of comorbid conditions impact surgical outcomes of older adults with cancer. The aim of our study is to evaluate the relationship between diabetes mellitus (DM), and preoperative HgbA1c (pHgbA1c) with frailty and adverse surgical outcomes among older adults with cancer. Methods: Retrospective study of patients aged 65 and older who underwent elective cancer surgery with at least one-day hospital stay in 2015-18 at our institution. For our primary aim, we assessed the association between DM and frailty, sociodemographic and clinical characteristics by reporting the characteristics stratified by presence of DM. Frailty was assessed using the Memorial Sloan Kettering Frailty Index (MSK-FI), which composes of 10 comorbidities and one functional component. This score ranges from 0-11, with higher score corresponding to greater frailty. Our secondary aim to assess the association between pHgbA1c and frailty was evaluated by a linear regression with pHgbA1c as the outcome and MSK-FI as the predictor. For our tertiary aim to assess the association between pHgbA1c and adverse surgical outcomes, we used a univariable logistic regression model with the outcome as the composite definition of 30-day readmission, 30-day emergency room visit, 30-day major complication, or 90-day mortality, and pHgbA1c as the predictor. We also used a multivariable logistics regression model, adjusting for age, gender, American Society of Anesthesiologist Performance Status (ASA-PS), and operation time.Analyses related to pHgbA1c excluded patients without pHgbA1c records. Results: Among 4417 patients, 1024 (23%) had DM and among those 403 (39%) had pHgbA1c values recorded. The results showed patients with DM were older, had higher ASA-PS, more comorbidities and higher MSK-FI scores than those without DM. Moreover, pHgbA1c was not found to be statistically associated with degree of frailty (β for one unit change in MSK-FI 0.11, 95% CI 0.00, 0.21; p = 0.057). Lastly, in both unadjusted and adjusted models pHgbA1c was not associated with adverse surgical outcomes (OR for both models, 0.99, 95% CI 0.83, 1.16, p = 0.9). Conclusions: In this cohort of patients whose pHgbA1c values were recorded, pHgbA1c was not associated with adverse surgical outcomes. Further studies are needed to confirm this study with more focus on perioperative care process of patients with DM and find cohorts whose pHgbA1c measurements can better predict frailty and adverse surgical outcomes.