Objective Patients with a small aortic annulus (≤21 mm) have an increased risk of patient-prosthesis mismatch after valve replacement. The aim of this study was to compare the early hemodynamic performance of the balloon-expandable transaortic valve implantation Edwards system (SAPIEN) and the sutureless Perceval prostheses. Methods Fifty patients underwent transcatheter aortic valve implantation, and 113 patients underwent sutureless aortic valve replacement. Mean ± SD aortic annulus diameter was 19.7 ± 1 mm, with no significant difference between groups. SAPIEN valve size was 23 mm in 40 patients (80%) and 26 mm in 10 patients (20%). Perceval valve size was small in 45 patients (40%), medium in 62 patients (55%), and large in 6 patients (5%). Transthoracic Doppler echocardiographic images were collected at baseline and before discharge. Results There were no significant difference in predischarge effective orifice area (SAPIEN: 1.5 ± 0.5 cm2 and Perceval: 1.48 ± 0.34 cm2, P = 0.58) and indexed effective orifice areas (SAPIEN: 0.93 ± 0.32 cm2/m2 and Perceval: 0.88 ± 0.22 cm2/m2, P = 0.42). Predischarge mean ± SD transaortic gradient was lower with the SAPIEN than with Perceval valves (12 ± 6 and 17 ± 6 mm Hg, respectively, P < 0.001). Rates of moderate and severe prosthesis-patient mismatch were similar (SAPIEN: 44% and 10% and Perceval: 50% and 14%, P = 0.53 and 0.75, respectively). There were no moderate-severe paravalvular leaks. Conclusions Although indexed effective orifice areas were similar, transcatheter aortic valve implantation with the balloon-expandable SAPIEN system yielded lower predischarge transaortic mean gradients than the surgically implanted Perceval, in patients with a small annulus.