Abstract Background Frailty increases the risk of post-operative complications in older patients, however clinical assessment of frailty is not routinely undertaken. We aimed to evaluate whether the Hospital Frailty Risk Score (HFRS), calculated using data available from ICD-10 coded electronic records at time of admission, could be used to identify patients at risk of complications. Methods Hospital and death data for individuals aged >50 years, admitted for surgery to all hospitals in NSW, Australia 2016-2017 were linked. Hospital-acquired complications (HACs) were identified using 16 categories. Logistic regression models were used to quantify influence of explanatory variables; age, sex, comorbidity, urgency of admission and HFRS on risk of HACs, and model performance assessed. Results Of the 150,355 hospitalisations, 12.3% resulted in a complication. Infections (6.1%) and delirium (3.3%) were the most common complications. 30-day mortality was higher following complication (12.5% vs 1.8%, p < 0.0001). Emergency surgery was the strongest predictor for >1 complication (OR 2.67, 95%CI 2.57-2.78), followed by multimorbidity (OR 1.90, 95%CI 1.78-2.02) and high HFRS (OR 1.57, 95%CI 1.48-1.66). The fully-fitted model provided acceptable discrimination (AUC 0.72, 95%CI 0.72-0.73). Conclusions The HFRS was strongly associated with complications following surgery. Used in conjunction with age, sex, urgency and comorbidity the HFRS predicted complications with 72% accuracy. Key messages This is the first study to validate the use of HFRS as a screening tool to identify frail patients at risk of developing complications following surgery. The HFRS provides hospitals with an inexpensive, systematic method to screen for high-risk patients at time of admission.