Objectives: To present the preliminary results of the IDEA project relating to the effect and impact of adverse events in two of the hospitals included in the study. Material and method: Design: prospective cohort study carried out in the first quarter of 2004 in two wards of two hospitals of the Autonomous Community of Valencia (Spain). Sampling: consecutive. Subjects: patients more than 14 years old, hospitalized for more than 24 hours in the selected wards. The total number of subjects was 673. Main measurements: proportion of adverse event alerts (positive detection questionnaire) and the accumulated incidence of identified adverse events (modular questionnaire for confirmation of AE and their characterization). Results: The incidence of alerts was 34.7% in hospital A (general surgery ward) and 31% in hospital B (internal medicine ward). The incidence of adverse events was 16.1% and 5.6% respectively. A total of 42.1% of the adverse events in hospital A and 16.7% of those in hospital B were avoidable, independently of the seriousness of their consequences. Conclusions: The adverse event surveillance systems differed between the two wards studied. Variability in clinical practice affects identification of adverse events on reviewing medical records.