OBJECTIVES: Many adults with repaired tetralogy of Fallot will require a pulmonary valve replacement (PVR), but there is no consensus on the best timing. In this study, we aim to evaluate the impact of age at PVR on outcomes. METHODS: This is a national multicentre retrospective study including all patients >15 years of age with repaired tetralogy of Fallot who underwent their first PVR between 2000 and 2013. The optimal age cut-off was identified using Cox regression and classification and regression tree analysis. RESULTS: A total of 707 patients were included, median age 26 (15-72) years. The mortality rate at 10 years after PVR was 4.2%, and the second PVR rate of 6.8%. Age at PVR of 35 years was identified as the optimal cut-off in relation to late mortality. Patients above 35 years of age had a 5.6 fold risk of death at 10 years compared with those with PVR under 35 years (10.4% vs 1.3%, P < 0.001), more concomitant tricuspid valve repair/replacement (15.1% vs 5.7%, P < 0.001) and surgical arrhythmia treatment (18.4% vs 5.9%, P < 0.001). In those under 50 years, there was an 8.7 fold risk of late death compared with the general population, higher for those with PVR after 35 than those with PVR below 35 years (hazard ratio 9.9 vs 7.4). CONCLUSIONS: Patients above 35 years of age with repaired tetralogy of Fallot have significantly worse mortality after PVR, compared with younger patients and a higher burden of mortality relative to the general population. This suggests that there are still cases where the timing of initial PVR is not optimal, warranting a re-evaluation of criteria for intervention.