During the MOREST project (2001–2006), oyster summer mortality was extensively investigated in Crassostrea gigas in France. However, significant increases in magnitude and geographic distribution of summer mortality episodes have been reported since 2008. This paper investigates the response to selection for survival, which was successfully performed between 2001 and 2003, in this new context. It also examines the relationship between the mortality and the three pathogens Vibrio aestuarianus, V. splendidus and the Ostreid herpesvirus 1 (OsHV-1). Three batches of juvenile (6-month-old) Crassostrea gigas were produced in February 2009 and tested in the field in the Marennes-Oléron Bay during summer (August and September) 2009: one control batch, one batch descended from one family selected to be resistant (R) to the summer mortality phenomenon and one descended from one family selected to be susceptible (S). For each batch, mortality was monitored and oysters were diagnosed for the presence and intensity of the three pathogens. A mortality episode started 11 days post deployment and had finished by day 17. A significant difference in mortality was found among the batches, with 5, 53 and 94 % for the R, control and S batches, respectively. This finding indicates that the selection previously made for resistance to summer mortality still confers an advantage for the survival of juvenile C. gigas to descendant batches in the context of the recent massive oyster mortality that has occurred in France since 2008. Disease diagnoses revealed V. splendidus at all sampling dates but never V. aesturianus, showing that only the first of these bacteria could have been involved in this mortality event. However correlations between the mortality and the presence or the bacterial load of V. splendidus were weak, negative and not significant. In contrast, high and significant correlations were found between the mortality and the prevalence of OsHV-1, as well as between the mortality and OsHV-1 viral load. This study clearly revealed the suddenness of the infection in juvenile C. gigas in field conditions, as no OsHV-1 was detected in seed at deployment, while all juvenile oysters tested 7 days post-deployment were infected and the peak of the viral load was observed 11 days post-deployment. Finally, this study is the first report of herpesvirus resistance in the R oysters. Even if all R oysters had been infected by the OsHV-1, they were able, first, to limit the amount of the viral load in their tissues, and, second, either to eliminate the virus from their tissue or to decrease the quantity of viral DNA to a level below the threshold of the real-time PCR technique, as OsHV-1 prevalence decreased from 100 % to 33 % for the R oysters screened at 7 and 17 days post-deployment, respectively.