Nervous necrosis necrosis virus (NNV) is the causative agent of viral nervous necrosis (VNN), one of the most serious diseases in over 30 species of cultured marine fishes worldwide. Although several kinds of NNV vaccines have been developed, none of these vaccines have been yet marketed. Here, we demonstrate the potentiality of a live NNV vaccine for sevenband grouper Epinephelus septemfasciatus at a low rearing temperature (17 degrees C). Moreover, we investigated the kinetics of NNV infectivity titer in fish reared at low and optimum temperatures (17 degrees C and 26 degrees C) for VNN onset to determine why sevenband grouper reared at 17 degrees C survive NNV infection. In pathogenicity tests of NNV, fish mortality was reduced by decreasing the fish rearing temperature, and no mortality was observed in fish reared at 17 degrees C regardless of the infection method. During fish acclimation to the optimum temperature of VNN onset (26 degrees C), increased mortalities were observed in the survivors from the 1st NNV-infection. Little or no mortality was observed in the 2nd NNV-infection. Thus, it was demonstrated that the survivors from the 1st NNV-infection mounted a specific protective immune response against NNV. Especially, in the fish infected with NNV by immersion at 17 degrees C, only two out of 30 fish died until the end of the 2nd infection (total survival rate: 93.3%), suggesting a positive potentiality for a live NNV vaccine. In the analysis of NNV kinetics in the fish reared at 26 degrees C, NNV rapidly multiplied up to >= 10(9) TCID50 before fish began to die, and the critical level of NNV was around 10(10) TCID50 g(-1). Probability of NNV multiplication reduced by decreasing the inoculated NNV dose, but NNV multiplication rate was independent of the NNV dose. The threshold of NNV for fish mounting a protective immune response was around >10(4) TCID50 g-1. Against this, in the fish reared at 17 C, NNV slowly multiplied in comparison with that in fish at 26 degrees C. NNV titer in the peak was at 10(7.1 +/- 1.4) TCID50 g(-1), which was far behind the critical level of NNV but still greatly above the threshold level (10(4) TCID50 g(-1)). Thus, it was demonstrated that the multiplication rate of NNV in vivo was strongly correlated to NNV virulence and fish mortality, and down-regulation of NNV multiplication in fish reared at 17 degrees C enabled control of VNN onset for development of a live NNV vaccine. (C) 2011 Elsevier Ltd. All rights reserved.