A mass mortality event of the red coral Corallium rubrum (L.) occurred in the NW Mediterranean region in summer 1999. The main objectives of the present study were to document the mortality suffered by the red coral populations in the Provence region and assess ecological correlates of the mortality which could help to identify the putative agent or agents of the event. The mortality outbreak resulted in partial to complete loss of the coenenchyme. The first observations of red coral mortality in the study area were in early October 1999 and continued until November 1999. To determine the extent of mortality on red coral populations, the incidence (= % colonies suffering mortality) and the virulence (= % coenenchyme loss) were quantified. Twelve surveys were carried out from November 1999 to January 2000. A total of 874 colonies of the 2375 observed showed some mortality. Incidence and virulence differed significantly among surveys. In some cases about 80% of the colonies were affected, while in others only 5% of colonies showed some mortality. Depth had a strong influence on the incidence of mortality. Shallower populations had the greatest incidence, while populations deeper than 30 m depth showed no mortality. However, virulence was similar in the 10 to 30 m depth range. With regard to habitat, the entrances of caves displayed a significantly higher incidence but not virulence than the interiors. Given the life history traits of red coral (mainly slow growth and low recruitment), it seems that recovery from the outbreak will be a long-term process. However, further studies are urgently required to provide basic information regarding red coral population dynamics as a basis for hypotheses on the actual recovery capability of affected populations. The cause of the mass mortality is unknown. However, during late summer 1999, the NW Mediterranean area affected by the mass mortality event experienced high temperatures and hydrographic stability over a period of several weeks. This temperature anomaly could have caused physiological stress or/and triggered the development of pathogenic agents that otherwise would have remained non-virulent. This hypothesis is reinforced by the absence of signs of mass mortality below 30 m depth in red coral populations. Since the NW Mediterranean seems to be affected by the global warming trend, if the temperature hypothesis is confirmed, the long-term consequences of a repetition of this kind of event could severely endanger the persistence of shallow-water red coral populations.