Large nationwide registries, such as those from Quebec, Denmark, and the Netherlands, are of paramount importance in the research domain of congenital heart disease (CHD) with relatively small patient numbers in compared with, for instance, coronary artery disease. In the interesting Danish registry-based study, Videbaek et al 1 showed that patients diagnosed with simple CHD in the 1960s have substantially increased long-term mortality and cardiac morbidity in comparison with the general population. In 1241 patients with simple CHD, the mortality rate was almost 2-fold higher than the Danish reference population. The unexpected increased mortality in these patients was mainly attributed to a 4-fold increased risk for sudden unexpected death. The authors conclude that further studies on the effectiveness of systematic medical follow-up programs appear warranted. As reported in the accompanying editorial, 2 these findings are highly relevant because current guidelines indicate no specialist care for many patients with simple CHD.Registry-based studies in CHD are warmly welcomed because large volumes of patients with CHD are needed to perform epidemiological studies. However, such registry-based studies also have several limitations. For instance, misclassification or underreporting of morbidity may occur because hospital records are not always systematically reviewed. As an example, the study by Videbaek et al reported only 15 registered cases of pulmonary arterial hypertension in a sample of 1100 patients at risk. This prevalence is remarkably lower than the 52 patients with pulmonary arterial hypertension in a similar population of 1115 patients with simple CHD at risk in our Dutch Concor registry.3 The possibility of a systematic review of clinical records for which patients provided informed consent in the Concor registry may explain the higher numbers in this registry.The Dutch Concor registry, however, has the limitation of incomplete inclusion. Because informed consent is needed for inclusion in this voluntary adult registry, adult CHD patients are not included automatically. Time-consuming identification of and approaching adult CHD patients account for the incompleteness of the Concor registry, whereas the voluntary aspect appears to be of minor importance, because 99% of patients identified are willing to participate in the Concor registry.Although the specific limitations of each large population registry may differ substantially from each other, registry-based studies may complement each other and similar results provide support for shared conclusions. The increased mortality in simple CHD lesions reported by Videbaek et al, for example, was also found in a recent study from the Concor registry. 4 Both study designs had important and different limitations, well described in the articles and accompanying editorials, 1,2,4,5 but eventually showed similar results. This is reassuring for the reliability of registry-based results.In summary, large registry-based studies are a must in CHD, because large pati...