AbstractBackgroundCongenital anomalies are a major cause of perinatal, neonatal and infant mortality.ObjectivesThe aim was to investigate temporal changes and geographical variation in survival of children with major congenital anomalies (CA) in different European areas.MethodsIn this population‐based linkage cohort study, 17 CA registries members of EUROCAT, the European network for the surveillance of CAs, successfully linked data on 115,219 live births with CAs to mortality records. Registries estimated Kaplan–Meier survival at 28 days and 5 years of age and fitted Cox's proportional hazards models comparing mortality at 1 year and 1–9 years of age for children born during 2005–2014 with those born during 1995–2004. The hazard ratios (HR) from each registry were combined centrally using a random‐effects model. The 5‐year survival conditional on having survived to 28 days of age was calculated.ResultsThe overall risk of death by 1 year of age for children born with any major CA in 2005–2014 decreased compared to 1995–2004 (HR 0.68, 95% confidence interval [CI] 0.53, 0.89). Survival at 5 years of age ranged between registries from 97.6% to 87.0%. The lowest survival was observed for the registry of OMNI‐Net (Ukraine) (87.0%, 95% CI 86.1, 87.9).ConclusionsSurvival of children with CAs improved for births in 2005–2014 compared with 1995–2004. The use of CA registry data linked to mortality data enables investigation of survival of children with CAs. Factors such as defining major CAs, proportion of terminations of pregnancy for foetal anomaly, source of mortality data and linkage methods are important to consider in the design of future studies and in the interpretation of the results on survival of children with CAs.