Abstract Introduction Prosthetic valve endocarditis (PVE) is the most unfavourable complication after any aortic valve replacement. Purpose The aim of this study was to analyse the impact of antibiotic-only and additional surgical therapy in patients with TAVI-PVE and to compare it with the results of patients with PVE after surgical aortic valve replacement (SAVR) regarding survival. Methods All patients who underwent primary isolated aortic valve replacement via SAVR (n=3.447) or TAVI (n=2.269) in our heart center between 01/2012–12/2018 were analysed. The diagnosis of PVE was made using the 2015 modified Duke criteria. Follow-up was performed until 06/2019 (max. 89, mean 34 months) with a follow-up rate of 98% regarding mortality. Results PVE incidence did not differ significantly between SAVR with 4.9 cases per 100 patient-years and TAVI with 2.4 cases per 100 patient-years (p=0.49), although TAVI patients were significantly older (mean 80 vs. 67 years) and had more multimorbidities (STS score mean 5.9 vs. 1.6) (p<0.001). TAVI prostheses with polymer showed a 4.3-fold higher risk to develop PVE than TAVI prostheses without polymer (p=0.004). Most common pathogens were staphylococci and enterococci in all PVEs. Antibiotic-only therapy in TAVI-PVE patients resulted in a significantly better 1-year survival compared with additional surgical therapy (90.9% vs. 33.3%; p=0.005) even in 4 patients with an indication for surgery according to ESC guidelines. In SAVR-PVE patients both therapies led to comparable survival (p=0.861). However, SAVR-PVE who were not operated although indication for surgery was given by the guidelines resulted in a significantly worse survival after 1 year (p=0.004). Conclusion TAVI patients did not have a significantly higher risk to develop PVE than SAVR patients despite their older age and greater comorbidities. The 1-year survival of TAVI-PVE patients was significantly higher after antibiotic-only therapy than after additional surgery, even if surgery had been indicated according to guidelines. This new finding is presumably due to the fact that TAVI prostheses do not have a polymeric suture ring. Our results show for the first time that indication for surgery in TAVI-PVE patients should be made carefully. SAVR-PVE patients however benefit from guideline-compliant surgery as in this respect antibiotics-only led to a significantly worse prognosis. This is probably related to the ability of gram-positive cocci to migrate into the polymeric ring and to protect themselves from antibiotic therapy by forming a biofilm. Some of the newer TAVI prostheses contain plastic components which can also be cured often with antibiotic-only but are associated with an increased risk to develop PVE. Funding Acknowledgement Type of funding sources: None.