In October 2003, a suspected massive influx of patients to the emergency departments in the northern health sector of Réunion Island was reported. Activity data and the characteristics of patients admitted to the emergency departments were collected and analyzed. Qualitative information was requested from the heads of the relevant departments and units. The circumstances surrounding the event were examined in the local press. A steady annual increase in activity at emergency departments was observed. In 2003, a slight increase in activity—which did not constitute a massive influx of patients—was noted during the first half of October. This fluctuation in activity cannot be attributed to a common environmental or infectious exposure. Regardless of activity levels, there is a structural problem with accommodating patients awaiting hospitalization. This study highlights some of the practical and methodological difficulties of using non-specific quantitative activity data in an alert system and confirms the value of the qualitative information provided by the network of emergency departments. (R.A.)