Transcatheter aortic valve-in-valve implantation (ViV) is a new treatment for failing bioprostheses (BP) in patients with high surgical risk. However, comparative data, using standard repeat surgical aortic valve replacement (redo-SAVR), are scarce. We compared outcomes after ViV with those after conventional redo-SAVR in two European centres with established interventional programmes. In-hospital databases were retrospectively screened for patients ≥60 years, treated for failing aortic BP. Cases of infective endocarditis or combined procedures were excluded. End-points were adjudicated according to the Valve Academic Research Consortium (VARC-2) criteria. From 2002 to 2015, 130 patients were treated (ViV: n = 71, redo-SAVR: n = 59). Age and logistic EuroSCORE I scores were higher with ViV (78.6 ± 7.5 vs 72.9 ± 6.6 years, P < 0.01; 25.1 ± 18.9 vs 16.8 ± 9.3%, P < 0.01). The 30-day mortality rate was not significantly different (4.2 and 5.1%, respectively) (P = 1.0). Device success was achieved in 52.1% (ViV) and 91.5% (P < 0.01). No stroke was observed after ViV but in 3.4% after redo-SAVR (P = 0.2). Intensive care stay was longer after redo-SAVR (3.4 ± 2.9 vs 2.0 ± 1.8 days, P < 0.01). Mean transvalvular gradients were higher post-ViV (19.7 ± 7.7 vs12.2 ± 5.7 mmHg, P < 0.01), whereas the rate of permanent pacemaker implantation was lower (9.9 vs 25.4%, P < 0.01). Survival rates at 90 and180 days were 94.2 and 92.3% vs 92.8 and 92.8% (P = 0.87), respectively. Despite a higher risk profile in the ViV group, early mortality rates were not different compared with those of surgery. Although ViV resulted in elevated transvalvular gradients and therefore a lower rate of device success, mortality rates were similar to those with redo-SAVR. At present, both techniques serve as complementary approaches, and allow individualized patient care with excellent outcomes.