Outcome studies in cardiovascular diseases such as hypertension are often based in a number of countries recruiting from multiple centres. This can lead to difficulties in end-point evaluation and classification. The INSIGHT study provided the unique opportunity to examine the role and usefulness of a Critical Event Committee (CEC) established to examine all events occurring in a cohort of 6321 patients and classifying them according to pre-determined criteria. More than 28% of investigator-coded primary events and more than 41% of secondary events were re-classified by the CEC. Particularly difficult diagnoses included myocardial infarction, angina and heart failure, and the database finally analysed was substantially altered after CEC Committee evaluation. Post hoc analysis of the INSIGHT database demonstrated that the primary events compared would have been reduced from 6.3% vs 5.8% of those randomized on nifedipine vs co-amilozide to 5.1% vs 4.7%--an overall reduction of 72 events of 382, a fall of 18.8%. We conclude that a CEC should be a crucial component of any large outcome study and that if all events are subjected to scrutiny, the quality of the database is greatly enhanced.