Syndromic surveillance emerged in the field of public health surveillance in the late 1990s. Initially proposed as a means of identifying acts of bioterrorism, it failed to convince many of its value or the benefits it could offer. The definition currently proposed by the Centers for Disease Control and Prevention (CDC) in Atlanta is the most widely accepted. It defines this surveillance as based on the automation of data recording and the transfer of recorded data for professional purposes. Of the many experiments and systems developed at that time, very few have survived. Only those systems that have integrated a public health approach through large-scale surveillance and strong involvement of official health surveillance agencies are still in use. Syndromic surveillance has many advantages, such as access to large amounts of data in real time, the absence of a specific workload for data entry, and the creation of useful historical databases for reference; however, its limitations must not be overlooked (sometimes limited sensitivity or specificity, significant technical constraints, etc.). Experience gained to date shows that syndromic surveillance should not be viewed as opposed to traditional surveillance. On the contrary, they should be viewed as complementary. Finally, the use of syndromic surveillance must be considered within a temporal framework: in the very short term for early warning, in the medium term for creating historical reference data, and in the long term for building historical databases useful for describing the health status of the population at the beginning of the 21st century. (R.A.)