OBJECTIVES More elderly patients (>80 years of age) are being referred for aortic valve replacement (AVR) with or without CABG. Current risk stratification models may not accurately predict the preoperative risk in these patients. We sought to determine which perioperative variables were relevant in determining short-term (30-day to in-hospital) outcomes in our intuition's series of consecutive AVR and AVR+CABG surgeries. We constructed a novel variable, patient-prosthesis mismatch (PPM) in the presence of diminished functional status (NYHA) classification, and studied its role as a predictor of mortality risk. METHODS From 2006 to 2010, 509 patients undergoing AVR or AVR+CABG were evaluated. We created four groups based on the age and procedure (AVR >80, AVR+CABG >80, AVR <80 and AVR+CABG <80). PPM was defined as a calculated effective orifice area index value of ≤ 0.85, and it was calculated from manufacturer-generated charts. In-hospital and 30-day outcomes were assessed using the Chi-square and logistic regression analyses. RESULTS Overall observed 30-day mortality for all groups was lower (n = 8, 1.6%) than the STS-predicted mortality. Reoperation and PPM+NYHA class III-IV were associated with short-term mortality, but age >80 years was not. Octogenarians referred for surgery often had advanced heart failure. CONCLUSIONS Overall, short-term outcomes after AVR with or without CABG were excellent and lower than predicted by the STS model. The low risk of AVR with CABG supports the consideration for earlier surgical referral and intervention for patients with a high likelihood of aortic stenosis progression before the onset of advanced heart failure ensues, regardless of the age. This should help further decrease the already very low mortality observed in these series. Efforts to avoid PPM in the setting of advanced heart failure may improve short-term results in this subset of patients.