Introduction - In July 2006, a major heat wave was observed in France. Methods - A model linking daily fluctuations in mortality and temperatures was developed based on data from the summers of 1975–2003. This model provides an estimate of the expected number of deaths on a given day, taking into account the temperatures observed on that day and the preceding 10 days. The observed and expected numbers of deaths in July 2006 were compared. Results - From July 11 to 28, 2006, excess mortality, estimated at approximately 2,100 deaths, was observed across France. This excess mortality is approximately 4,400 deaths lower than what would have been expected if the population’s vulnerability to heat waves had remained the same as during the 1975–2003 period. Discussion and Conclusion - The lower-than-expected mortality during the July 2006 heat wave can be interpreted as the result of a reduction in the population’s vulnerability to summer heat waves. This reduced vulnerability of the population could be the result of developments following the August 2003 heat wave: general awareness of the risks, implementation of preventive measures, and the establishment of a heat wave monitoring and warning system. (R.A.)